Introduction
The prevalence of psychiatric illness in the general population is 30%.1 Of this population, 2% could have depression, 10% with anxiety disorders and 8% mixed symptoms of depression and anxiety.2-4 In primary care (PC) it is estimated that 20% of the patient population could present with a depressive disorder and that 30% of the psychiatric visits in PC are secondary to symptoms of generalised anxiety.5,6 There is an under-diagnosis of depressive illnesses and often the patients diagnosed and treated do not receive sufficient doses and treatment times.5
An increase in the use of antidepressants (AD) in Spain has been reported in recent years, particularly due to the use of selective serotonin reuptake inhibitors (SSRI). This fact could be due to the following factors: a) under use of the classic AD associated with the high incidence of secondary effects or their potential lethality in cases of overdose; b) the better safety profile of SSRI; and c) a widening of the indications for AD, as well as using this group of drugs for the treatment of poorly defined symptoms.4,5,7,8
At the same time as the increase in the use of AD, an increase in the use of benzodiazepines is also detected, particularly those with a short half-life, which is not in keeping with the trends in other European countries.9-11
In Catalonia, the obtaining of data on the use in defined daily doses/1000 inhabitants/day (DDD) has provided evidence of an increase use of BZD and AD in recent years.
The objective of this study is to get to know the profile of use of AD and tranquilizer-hypnotics (T-H) in PC in Catalonia and find areas for improvement.
Patients and Methods
A transversal, observational, epidemiological study of prescription-indication has been carried out in the primary care services (PCS) of Baix Llobregat Litoral (BLL) and Mataro-Marense (MM) of the Catalonian Institute of Health, during the months of February and March 2002. The doctors from these 2 PCS were requested to include, by consecutive sampling, 10 patients who attended the clinic and who were on treatment with AD and/or AH. The participation by the doctors was voluntary and the analysis of the responses was anonymous.
The selection of the sample of doctors was different in the 2 PCS. In the BLL PCS, all the doctors in the area were included (n=133) and in the MM PCS a simple random sample of 42 doctors out of a total of 113 were selected, calculated from the DDD of fluoxetine in 2001.
The AD were classified as: classic, second generation, SSRI, and other AD, and the T-H as: short half life of elimination (t1/2) BZD, long t1/2 BZD, and other tranquilizers.
The variables analysed were: age and sex of the patients, type and number of AD and T-H prescribed, treatment scheme, main reason for the prescription, and the professional who initiated it. To contrast the validity of the sample with the reference population, the uses of AD and T-H (percentages of DDD from the same period as the study) of each PCS were compared with those obtained in the study sample.
A descriptive analysis of the profile of use of the different psychiatric drugs and their distribution according to age and sex was carried out. For the comparisons the *2, the Student t test, and analysis of variance was used, depending on the type of variable. Statistical significance was established at 5%.
Results
A 58.3% response (102 professionals) was obtained, with a total of 998 patients included. 1613 drugs were evaluated, of which 43% (n=693) were AD and 57% (n=920) T-H.
In Figure it is seen that the profile of use of the different active principles in the sample studied was similar to the data of use (DDD) obtained in the 2 PCS during the period of the study, except for lormetazepam and sertraline (less in the sample) and dipotassium clorazepate (more in the sample).
Patient Characteristics
In Table 1 the comparative demographic characteristics between the 2 PCS of BLL and MM are shown. The majority were women (76%). The mean age was 52.2 years (range, 16-96) and was higher in the MM PCS.
The mean of drugs per patient was 1.6±0.71. 50.2% (n=501) of the patients only took one drug, 39.2% (n=391) 2, 9.2% (n=92) 3, and 1.4% (n=14) took 4.
The prescribing of AD was higher in women (n=509; 67.7%) than in men (n=142; 59.9%) (P=.028). No differences were observed in the T-H according to sex. As can be seen in Table 2, 45.8% of the cases received combined treatment, which was higher in women (48.3%) than in men (38%) (P=.019).
The patients who only received one tranquilizer treatment had a higher mean age (56.6±16.7 years) than those who took only AD (50.2±17.6 years), or combined treatment (49.9±16.1 years) (P<.001).
On analysing the data according to PCS, treatment with only AD was higher in the BLL PCS, while treatment with only T-H was higher in the MM PCS. This difference was significantly different (Table 2). No differences were found in combined treatment between the PCS.
Treatment Characteristics
The most prescribed therapeutic sub-groups were the SSRI (31.6%), the short t1/2 BZD (32.9%), and the long t1/2 BZD (19.8%), with no differences between the PCS (Table 2).
The AD most used were fluoxetine and paroxetine. The most prescribed T-H were alprazolam, lorazepam, diazepam, and dipotassium clorazepate (Table 3).
The combinations most used were: fluoxetine and paroxetine with alprazolam (n=47 and n=45, respectively) or lorazepam (n=25 and n=21, respectively).
The AD were indicated by the PC doctor in 45.7% and by the PC psychiatrist in 33.1% (Table 4).
The professional who initiated the prescribing of AD was the PC doctor on more occasions in the BLL PCS compared to MM (49.7% as opposed to 36.3%; P=.04). The prescribing of SSRI was carried out by the PC doctor in 55.3% of cases, while the prescribing of classic AD and
other AD were initiated in the majority of occasions by the PC psychiatrist (36.6% and 60%, respectively).
The T-H were initiated in 49.1% by the PC doctor and 30% by the PC psychiatrist (Table 4). By sub-group, the PC doctor was the one who initiated the majority of treatments with short t1/2 BZD in 49.6%, long t1/2 BZD in 46.2% and other tranquilizers in 58.2% of cases.
The main reasons for prescribing were chronic depression-dysthymia (35.8%) and major depressive episodes (24.4%) for AD and generalised anxiety (33.3%) and insomnia (23.9%) for the T-H (Table 5).
On further analysing the reason for prescribing according to therapeutic subgroup it can be seen that the major depressive episode along with chronic depression-dysthymia were the main reasons for prescribing SSRI, as well as the AD and T-H (61.2%, 49.3%, and 63.5%, respectively). As regards the T-H it is observed that generalised anxiety is the main reason for prescribing long t1/2 as well as short t1/2 BDZ (41.3% and 31.6%, respectively). The main reason for prescribing the other BZD was insomnia (60.9%).
40.1% (n=268) of the AD and 51.5% (n=441) of the T-H had been prescribed for more than 1 year. No differences in the length of treatments were found between PCS.
As regards the AD, when the prescription was started by the PC doctor, the duration of treatment at the time of collecting the data was <3 months in 49.2% (n=148) and >1 year in only 22.6% (n=68). If the prescription was initiated by the PC psychiatrist, in 14.7% (n=32) it was <3 months and in 56% (n=122) >1 year.
As for the T-H indicated by the PC doctor, 39.3% (n=164) had been prescribed for up to 3 months and 37.6% (n=157) for more than 1 year. Of those prescribed by the PC psychiatrist. 12.8% (n=32) had a duration of <3 months and 67.6% (n=169) >1 year.
Fixed doses were used in 99.1% (n=659) of the AD and in 81.0% (n=715) of the T-H, no differences being detected as regards the indication. 99.3% (n=434) of the combined treatments were also prescribed in fixed doses.
Discussion
In the present study, the population characteristics, as well as the profile of the use of psychiatric drugs, are similar in the 2 geographic areas studied. The population on treatment with psychiatric drugs is middle-aged, mainly female and the majority of patients receive treatment with T-H only, or combined with AD. The age of the patients receiving treatment with T-H is higher than those who receive AD.
The larger percentage of women being treated agrees with results from other published studies.12-14
The possible causes of this could be a higher prevalence of psychiatric illnesses in the female population or a greater frequency of visits to the PC centres.13 The greater use of T-H found, as compared to AD, has also been reported in similar populations. In a study on the use of psychiatric drugs in PC12 a greater use of BZF as compared to AD was detected, a difference which was higher than the data obtained in the present study.
It is observed that a larger number of prescriptions originate from the PC doctor than the other specialists, although this is lower than that recorded in other studies (56% and 68%).12,13
The AD most prescribed are the SSRI. The majority of its prescribing has been initiated by the PC doctor, while the psychiatrist initiated the prescription of the classic AD and other AD. This profile of use is similar to that reported in another study.15 In our study, the duration of treatment is prolonged. More than 50% of T-H treatments were prescribed more than 1 year ago, and 56.6% of the AD more than 6 months ago. Prolonged treatment with BZD would not be in accordance with the recommendation that it should not be taken for periods >3 month, giving that long-term use causes tolerance and dependency.16 In our study it is noted that the T-H established by the psychiatrist were maintained for more prolonged periods than those initiated by the PC doctor. A possible explanation for the difference could be that the patients referred to the specialist could have a more serious illness, or also that the PC doctor does not normally modify or withdraw medication initiated by other specialists.
As regards the treatment with AD, the results are surprising if the indications for which the majority are prescribed are considered (major depression and chronic depression) and the recommendations of the duration of treatment of 6 months for a first episode and 5 years in a second episode. The longer duration of a treatment, when this has been prescribed by the psychiatrist could be due to the characteristic differentials in the sub-group of patients referred to the specialist or a result of the design of the study itself (possible selection bias by the use of prevalent cases). There are few studies on the induction of psychiatric treatments in PC. In 1 study,17 it is reported that the prescribing of BZD is common among PC doctors and is little influenced by the specialist, results which disagree, in part, with those obtained in our analysis. Similar studies with AD have not been found.
The high proportion of patients on combined treatment is emphasised, much higher than the 25.8% found in another study.12 On the other hand, in another study carried out to determine the pattern of use of BZD and AD in the elderly, a combined use (50%) similar to that reported in our study was found.18 Combined treatment could be justified at the beginning of the depression but, once the latency period of the AD has passed (4-6 weeks), it would be expected to withdraw the T-H. Therefore, the high percentage of combined treatments, which could correspond to inappropriate maintenance treatment with T-H, is surprising.
This study has certain limitations, such as the possible bias in the participation of the doctors, by being voluntary. It can have a possible bias on the selection of subjects by having selected those who had a prescription similar to that recommended in PC or in those who might have had a higher initiation of prescription by the specialist. In this sense, the similarity obtained between the profile of the active principles of our sample and those prescribed in our areas would be in favour of the representivity of the sample.
The transversal design itself limits the conclusion as regards the variables such as treatment times, which could be underestimated or overestimated. The shorter duration of treatment with AD prescribed by the PC doctor could be due to the higher frequency of those patients visiting the health centres at the start of treatment.
The high use of T-H, as well as the profile of use of AD and T-H detected in our population, suggests the need to develop strategies in our PC centres to reconsider the prescribing of psychiatric drugs, in terms of reducing the prescription and duration of the different T-H, as well as to improve the coordination between the PC psychiatrists and doctors. In this sense, a line of work could be the production of a performance and referrals protocol agreed between the different professionals involved in the problem of the various mental illnesses.
Acknowledgements
To all the PC doctors of the Baix Llobregat Litoral and Mataro-Marense of the Catalonian Institute of Health who, in an unselfish way, collaborated in the collection of data.
