Abstract
BACKGROUND:
Within recent years, the increasing popularity of complementary and alternative medicine (CAM) has led health care authorities to focus on the safety of these drugs. One reason for the low awareness of adverse drug reactions (ADRs) associated with CAM might be that users and physicians believe that there are no risks associated with CAM drugs. Recent studies have shown that ADRs are under-reported and are considered a leading cause of morbidity and mortality. The Evaluation of Anthroposophical Medicine (EvaMed) Pharmacovigilance Network was formed in 2004 at the Havelhoehe Research Institute in Berlin and is composed of 38 CAM physicians located in 12 of the 16 federal states in Germany for the purpose of using EvaMed data to evaluate the prescribing patterns, effectiveness, and safety of CAM therapies.
OBJECTIVES:
To describe and quantify the volume and severity of ADRs for CAM and conventional (CON) drugs in a proprietary database created from prescriptions and patient data of primary care CAM physicians who participate in the EvaMed Network.
METHODS:
This was a prospective, multicenter, observational study based on the ADR reports and electronic prescription data of 38 individual physicians (21 general practitioners, 9 pediatricians, 4 internists, 2 gynecologists, 1 dermatologist, and 1 neurologist) participating in the EvaMed Network. In addition to standard medical education, all physicians had 5 years practical experience and an additional qualification for anthroposophic medicine, which is a subcategory of CAM. All 38 physicians documented ADRs deemed serious, defined as life threatening or resulting in death, disability/incapacity, or inpatient hospital days. Due to the time-consuming nature of documenting ADRs, only a subgroup of 7 physicians (4 in general practice and 1 each in internal medicine, pediatrics, and gynecology) agreed to report both nonserious and serious ADRs. Therefore, the incidence and frequency of ADRs were evaluated in this subgroup. The study period was January 2004 through June 2009. ADRs were documented by the physicians using an electronic case report form in the EvaMed software, which was linked to the physicians’ existing electronic medical record (EMR) systems and incorporated into their daily routines to avoid missing data or double entries. The participating physicians were compensated €15 (approximately US$20) for each ADR report. All ADR reports were monitored at the Havelhoehe Research Institute by 2 physicians who evaluated patient characteristics, present visit diagnosis, target drugs, associated drug classes and type of drugs, type of ADR, actions taken for the ADR, and outcome of the ADR.
RESULTS:
There were 1,018,626 drugs (54.8% CAM) prescribed by the 38 physicians for 88,431 patients, and 412 ADRs reported for 389 patients; 124 (30.1%) ADRs were for CAM drugs. The majority were reported in children (69.2%, n = 285) and females (56.3%, n = 232). All serious ADRs (n = 14) were associated with CON drugs. In the subgroup of 7 physicians who agreed to report all ADRs, a total of 327 serious and nonserious ADRs were reported for 392,243 prescribed drugs (0.08%) and for 308 of 25,966 patients (1.2%). ADRs were reported for 241 of 16,032 children (aged 17 years or younger; 1.5%) versus 67 of 9,934 adults (0.7%). Of the 327 total ADRs, 10 (3.1%) were serious. There were 95 ADRs for 213,900 CAM prescriptions (4.4 per 10,000) versus 232 for 178,343 CON prescriptions (13.0 per 10,000). The CAM drug with the highest frequency of ADRs was Pelargonium sidoides root (0.21%, 4 of 1,940 prescriptions). The most frequently reported ingredient in CAM was ivy leaves with an ADR frequency of 0.17% (n = 11 of 6,575 prescriptions). The most reported drug connected with ADRs was amoxicillin (1.36%, n = 31 of 2,276 prescriptions). The most common ADR medical management was withdrawal of the drug (82.3% overall, 83.9% CAM, 81.6% CON).
CONCLUSIONS:
A sample of 38 CAM physicians reported the occurrence of at least 1 ADR for 0.4% of treated patients in a 5.5-year study period. There were no serious ADRs reported for CAM drugs. In a subsample of 7 physicians who agreed to report all nonserious and serious ADRs, 1.2% of patients experienced at least 1 ADR; rates of ADRs per 10,000 prescriptions were 4.4 for CAM drugs and 13.0 for CON drugs.