Patients
The study was conducted in medical departments for adult patients of the governmental University Medical Centre Ljubljana (UMCL), the primary city hospital for the Slovenian capital city of Ljubljana serving a population of 400,000 inhabitants and tertiary referral hospital centre serving a population of 2,000,000. It includes nine specialised medical departments with 82 regular beds dedicated to gastroenterology, 74 beds cardiology, 50 beds angiology, 42 beds endocrinology, 32 beds nephrology, 26 beds haematology, 14 beds intensive care, 13 beds pulmonology, and 6 beds to toxicology. Patients with solid tumours are treated mainly at the Institute of Oncology and were not included in the study. The patients are admitted to medical departments through a medical emergency department that covers all medical departments, through outpatient departments of individual specialised medical departments and those transferred from other primary hospitals and non-medical departments of UMCL, or whose admittance was planned for special diagnostic procedures or therapy.
Date collection
In the study a team of three internal medicine specialists, of whom one was a postgraduate in clinical pharmacology, retrospectively reviewed 520 complete medical records with all patient medical documents chosen at random by a computer for ADRs causing admissions. The number of included patients was computed according to the main aim of the study, which was the evaluation of ADR related admissions depending on the method of reporting and detection (3 possible answers), urgency of admissions (4 possible answers) and medical departments (9 possible answers). 16 possible answers regarding the method of reporting and detection, path of admission and medical department were multiplied by 30 cases per each answer, giving us 480 patients. Finally, 480 patients were rounded up to 520, which represents 3% of 17,500 patients hospitalised in medical departments in UMCL in 2006.
The definition of ADRs used in the study was that of the WHO: "Any noxious and unintended response to a drug that occurs at doses normally used in humans for the prophylaxis, diagnosis or therapy of disease" [
10]. The assessment of ADR causality was performed also using the WHO system [
11]. The study team had to reach a consensus decision on the certainty of ADR related admission and excluded doubtful cases. ADR causing admissions were considered as certain if ADR had a plausible time relationship to drug intake, plausible response to withdrawal, definitive pharmacological or phenomenological explanation and could not be explained by disease or other drugs [
11]. Patients with ADR on admission who were not directly responsible as well as those who received a deliberate or unintentional overdose or relapsed because of non-compliance were not considered to have ADR related admission. Afterwards all certain cases of ADR related admissions regarding WHO causality criteria were classified according to ADR type. The type of ADR estimated its preventability and suggested preventive measures for each ADR such as closer patient clinical and laboratory monitoring, applying protective measures and other selection of drugs. These were checked for their feasibility in every patient individually by the study team revising patient medical documentation.
For all patients age, sex, number of prescribed drugs on admission, discharge diagnosis with their ICD-10 codes performed by treating physicians, alcohol abuse, renal failure, liver failure, urgency and path of admittance and subspecialty of medical department were recorded. Renal failure was considered in patients with serum creatinine levels 50% above the upper normal level (150 micromol/L) and liver failure in patients with liver enzymes (alanine aminotransferase, aspart aminotransferase and gamma glutamyltransferase) at twice the upper normal level.
ADR causing admissions, documented and described in the medical records by the treating physicians who were unaware of the study and not specially trained and encouraged to record, code or report ADRs, were carefully recorded and additionally reassessed by the study team, who met to reach a consensus decision on the certainty of ADR related admission according to WHO causality criteria documented by the treating physicians and excluded doubtful cases.
Subsequently the hospital information system based on ICD-10 coding was checked for whether ADR related diagnoses causing admission of included patients were properly coded. ICD-10 codes including the words "drug induced" or "due to" a medicine or which are recognised to be invariably caused by a drug along with external cause codes indicating that a drug was implicated were considered as an ADR related diagnosis. The results of ADR coding are expressed as the percentage of ADR related admissions documented by the treating physician that were not properly coded and the percentage of ADR related admissions in patients included in this study that would be established only by searching a hospital information system based on ICD-10 coding.
Finally, the Slovenian national spontaneous ADR reporting system database was searched for patients with ADR reported by UMCL in 2006. The Slovenian National Pharmacovigilance Centre was established in 1984 and has been a member of Eudravigilance since 2004. The treating physicians should report all ADRs to the national reporting system using the ADR reporting form, but reporting is spontaneous and no repercussion is followed for not reporting. In 2006 the National Pharmacovigilance Centre received 100 ADR reports per one million inhabitants, mainly from hospital physicians. 20 ADR reports were sent to the National Pharmacovigilance Centre from UMCL in 2006. The patients in the national database of reported ADRs and patients included in this study were matched according to date, reporting hospital, patient's initials, age, drugs and reported ADRs. If all matched well, we assumed this to be the same case. The results of reporting ADR related admissions to the national reporting system are expressed as a percentage of ADRs described by the treating physician in medical records that were not reported to the national spontaneous reporting system and a percentage of ADR related admissions in patients included in this study that would be established only by searching the national spontaneous reporting system database.
A Slovenian National Medical Ethics Committee approved this study.