Abstract
Introduction
Following the Agence Française de Sécurité Sanitaire des Produits de Santé (Afssaps) (French Health Authority) recommendations in 2001, which impose the rigorous follow-up (electrocardiogram [ECG] and ionogram) of patients treated with antipsychotics (AP), a monitoring protocol was elaborated and set up in the Caen psychiatric hospital in April 2002. Protocol evaluation compared with fixed aims was performed after two years' follow-up.Aim of the study
This protocol had to answer a triple aim: better identification of patients at risk, ensure in-treatment monitoring, be simple and adapted to daily practice.Inclusion criteria
A systematic admission check-up (S0) which includes cardiological and biological controls and after one and six months' treatment control (S1 and S6) were recommended. The major risk factors (RF) researched were long QT interval, bradycardia and hypokaliemia.Results
The initial monitoring was conducted in 601 patients (that only corresponded to 17% of hospital admission active files during the considered period). Means delays before obtaining an ECG were three times those obtained existing biological check-ups (11 days versus three days after date of admission). Systematic and integrated characterisation controls on admission were not respected. We noted that two-third of patients admitted during this period were hospitalized for only five days, although the mean time to obtain an ECG is of 11 days. This delay (approximately one week) between ECG and biological check-up is not compatible with a complete patient RF evaluation. Respectively, 83 and 68 patients were controlled under treatment at S1 and S6. Only half of the patients were controlled at the one-month (S1) and 16% at the six-months' theoretical dates (S6). These delays are inappropriate, notably with regard to the mean time of hospitalisation (15-17 days). The incidence of major RF was higher in treated (71%) than in non-treated patients. Major RF presence at S0 was not systematically associated with an AP treatment contraindication. The excessive delay before the first ECG could partially explain why this initial check-up was not able to detect a pretreatment contraindication. On the other hand, AP treated patients who presented at least one major RF at S0 were more frequently monitored at S1 than patients who did not (26% versus 13%, p=0.05). Among the 168 patients treated with AP or other drugs prolonging the QT interval at risk, 33 had at least one follow-up. This risk population was not better controlled than the initial cohort.Discussion
Protocol evaluation is essential to improve its interest and feasibility. If systematic characterisation is simple, its application in practice is very difficult. The second version of the protocol presented here proposes to substitute the systematic ECG characterisation with the classification on admission of patients in "risk groups" that will condition the subsequent monitoring. Risk groups are identified into two RF types: those which are not related to AP treatment and those which are therapeutic attitude is adapted according to initial QTcorrigé (QTc), its progression between S0 and S (seven days) and kaliemia.