Frequency and Severity of Adverse Drug Reactions Due to Self-Medication: A Cross-Sectional Multicentre Survey in Emergency Departments

Approvals

The study protocol and patient informed consent procedures were approved by the Ethics Committee (St. Etienne CHU on 10 February 2008), and the Committee on Information in Health Research (CCTIRS/CNIL), according to French rules in clinical research.

Sampling and Randomization

A high volume of visits in participating EDs precluded uninterrupted prospective screening for inclusion throughout the study period. Additionally, as rates of hourly ED visits varied markedly within each day and from one day to another, we defined 13 time slots a priori covering the 24-h day as follows: 10 time slots of 1 h (from 10:00 am to 2:00 pm and from 5:00 pm to 11:00 pm), one time slot of 8 hours (from 11:00 pm to 7:00 am) and two time slots of 3 h (from 7:00 am to 10:00 am and from 2:00 pm to 5:00 pm). Subsequently, we randomly allocated these 13 predefined time slots throughout the 8 weeks of the study period for each participating ED. Randomization was done with computer-generated codes prior to the study enrolment period by our clinical research unit, which was not involved in data collection or patient care. Allocations were disclosed to research staff in every participating ED just prior to the study enrolment period. This method was designed to limit the potential for sampling bias. Additionally, patient demographics (age, gender and acute severity triage score) [9] were collected from administrative data of each participating centre over the same 8-week enrolment period. These data were compared with the overall study population to verify the representativeness of the ED population studied.

Patient Enrolment

All adult patients presenting to participating EDs during one of the predefined time slots were eligible for study enrolment. On entry, they were informed with a specific form about the study and the opportunity to participate. Medical or pharmacy students (hereafter designated as research staff) in every participating ED were specifically trained to screen candidates for study enrolment, using standardized screening forms. Consenting patients were subsequently included in the absence of exclusion criteria.

Exclusion Criteria

The following were precluded because we aimed to describe self-medication behaviour and unintentional related ADRs: (i) patients unable to participate because of cognitive impairment, neuropsychiatric disorders, language barriers or having presented with an unstable medical illness in the absence of a near relative who could answer for them; (ii) patients presenting for attempting suicide; and (iii) declining study participate (a written information form was submitted to patients and/or their relatives at the time of their admission to the ED). In each instance, the reason for exclusion was systematically recorded.

Variables and Data Collection

Self-medication behaviours were explored by a standardized questionnaire that had been previously built, implemented and tested in one centre [10] (see electronic supplementary material). This questionnaire is divided into two parts. The first part consists of a set of 20 closed-ended questions exploring all indications and dimensions of self-medication. The second part collects the characteristics of each medication cited by the patient during the first part (dosage, time between last dose and the ED visit, origin).

The method of data collection during the ED evaluation was then tested in three voluntary centres, which included standardized interviews of patients and/or their surrogates, as well as review of the medical record (i.e. physician notes and orders, laboratory reports, nursing notes, discharge instructions and ongoing prescriptions). Special attention was paid to all medications taken within 2 weeks prior to patient enrolment, including prescribed and non-prescribed drugs. All data were entered using online electronic case report forms (e-CRF), which allowed for real-time assessment of data completeness and patient follow-up. Data collection was performed by the local research staff, which was monitored by a clinical research pharmacist and supervised by the investigators.

Adverse Drug Reaction (ADR) Identification Process

The primary outcome was the diagnosis of ADR-SM and the identification of clinical and biological findings related to the effect of the drug(s). The investigators reviewed all cases to identify ADRs in each study centre, based on VIDAL dictionary (French book summarizing the characteristics of all medications, including pharmacology, adverse effects and drug-drug interactions). The local investigators were helped by the Naranjo scale [11] for drug causality assessment. Nevertheless, whatever could be the result of this score, they were asked to transmit all clinically relevant cases. The severity of the ADRs was assessed according to the Common Terminology Criteria for Adverse Events (CTCAE) [12], as (A) spontaneous regression; (B) regression after symptomatic treatment; (C) hospitalization with no life-threat; (D) life-threatening risk; and (E) death. The diagnosis of ADR and the drug causality assessment were then documented in the e-CRF. If necessary, notification of cases of drug toxicity was provided to the pharmacovigilance regional centre at the discretion of the local investigator.

All contentious issues transmitted by local investigators, every ADR-SM case (whether contentious or not), and some randomly assigned files were reviewed by an expert committee comprised of therapeutics professors, clinical pharmacists and emergency physicians, whose meetings and minutes were managed by the clinical research pharmacist. Furthermore, the entire database was scrutinized by the clinical research pharmacist in order to detect each case potentially related to an ADR; the expert committee was asked to assess such cases and to confirm drug causality (in order to validate the main outcome). Every local investigator was also asked to verify each subject file and to transmit all useful information regarding the possibility of an ADR to the expert committee. This committee was finally able to resolve each contentious case, and to validate the entire database. Last, the expert committee determined, for each ADR-SM case, the type of self-medication leading to the adverse event: self-modification of a prescribed treatment, discontinuation of treatment, non-prescribed drugs, or a drug interaction with non-prescribed drugs (i.e. self-prescription).

Grouping of Data

The diagnosis of the chief complaint and that of the ADR were first encoded to the International Classification of Diseases, 10th revision (ICD-10). To further improve grouping, data were re-coded using a standardized classification designed by the Société Française de Médecine d’Urgence (SFMU: French Society of Emergency Medicine) [13]. All drugs cited by patients, whether prescribed or not, were encoded to the Anatomical Therapeutic Chemical (ATC) classification system [14]. Medications not covered by the ATC were encoded as ‘Z’ (herbal medicines, vitamins, food supplements, calcium, magnesium, diosmine, anti-nausea or anti-diarrhoea pills, some medicines for constipation, balms and topical emollients, topical medicines for common cold, some anti-tussive syrups, omega 3 …).

Pharmaceutical Data

Of the patients included, 59.9 % took at least one prescribed medication, and 63.7 % self-medicated during the previous 2 weeks. Additionally, 84.1 % declared a self-medication behaviour (Table 2). Of the 11,724 drugs taken by the study population, 32.5 % were in a self-medication manner, and the most frequent were analgesics (n = 2,184, 75 % self-medication). Among the 3,848 drugs used in self-medication, origin was most frequently a non-prescribed medication purchased over the counter (OTC) at the pharmacy (50.5 %), followed by the rest of a previous prescription (19.9 %) and the use of a conditional prescription (14.5 %). Less frequent sources of self-medication were drugs supplied by relatives (5.3 %), the use of a prescribed drug with self-modification of the dose or duration (2.1 %), drugs purchased by mail or internet (1.2 %) and other unspecified sources (6.5 %).

Frequency of ADRs Related to Self-Medication (ADR-SM)

Of the entire cohort, 9.8 % (296/3,027) of patients experienced an ADR that was related to self-medication in 52 cases (Fig. 1). Depending on the population considered for the denominator, the rate of ADR-SM could be expressed as 17.6 % of patients experiencing an ADR (52/296), 1.7 % of the study population (52/3,027) or 2 % of patients reporting a self-medicating behaviour (52/2,556).

• ADR-SM related to prescribed drugs (self-medication behaviour): The type of self-medication leading to an ADR was most frequently associated with prescribed drugs, as a self-modification of a prescribed treatment in 21 cases or discontinuation of treatment in 14 cases. Finally, about two-thirds of ADR-SM are subsequent to the patients’ own decision on prescribed treatment.
• ADR-SM related to non-prescribed drugs: The use of non-prescribed drugs occurred in 16 cases of ADR-SM, and drug interaction with non-prescribed drugs in 1 case (so non-prescribed drugs led to a total of 17 cases of ADR-SM). The rate of ADRs related to non-prescribed drugs was 32.7 % of ADR-SM, 5.7 % of ADRs, and approximately 0.9 % among patients taking non-prescribed drugs during the previous 2 weeks (17/1,927).

The authors thank the MEDQUAL office for its logistic and technical support, Bioscience Writers, LLC, for editing support and Dr Hervé Maisonneuve for his contribution to the manuscript. We are also grateful to the National Medicine Academy and the French Society for Clinical Pharmacy for their support, to the French Association of Therapeutics’ Professors (APNET) for promoting this study, and to all members of the collaborative study group.

This article was published in a supplement sponsored by the Foundation for the National Institutes of Health (FNIH). The supplement was guest edited by Stephen J.W. Evans. It was peer reviewed by Olaf H. Klungel who received a small honorarium to cover out-of-pocket expenses. S.J.W.E has received travel funding from the FNIH to travel to the OMOP symposium and received a fee from FNIH for the review of a protocol for OMOP. O.H.K has received funding for the IMI-PROTECT project from the Innovative Medicines Initiative Joint Undertaking (http://www.imi.europa.eu) under Grant Agreement no 115004, resources of which are composed of financial contribution from the European Union’s Seventh Framework Programme (FP7/2007-2013) and EFPIA companies’ in kind contribution.