As evidence accrues of the efficacy of fluoxetine in treating depression in adolescents, US investigators report the impact of drug therapy with or without cognitive behavioural therapy (CBT) for continuation and maintenance treatment (Arch Gen Psychiatry 2008 65:44755).

Originally, 439 adolescents aged 12–17 years were randomised in a trial of 12 weeks' treatment with fluoxetine, CBT, combination therapy or placebo. For this study, those who had initially received placebo were excluded and, of the 327 initially assigned to active treatment, 270 completed the 12week trial. Of this group, 242 continued their assigned treatment for a further six weeks before an additional 18 weeks' maintenance treatment with CBT and/or fluoxetine.

Over the first 12 weeks, response rates were 42 per cent with CBT, 68 per cent with fluoxetine and 71 per cent with combined treatment. Eighty‐two per cent of those with an initial response continued and maintained their response for 36 weeks; success rates were 97 per cent (CBT), 74 per cent (fluoxetine) and 88 per cent (combination). Of those who did not attain a response initially, 62–80 per cent did so by week 36 with no significant differences between treatments.

The authors conclude that most adolescents will achieve a sustained response eventually and highlight the finding that CBT may help others to maintain their early response.

Three studies have reported new findings on the safety of psychotropic drugs during pregnancy.

Canadian epidemiologists describe a case‐control study to explore the effects of duration of antidepressant therapy during the first trimester (Br J Psychiatry 2008;192:344–50). They identified 2329 women who had been pregnant between 1998 and 2002 with a psychiatric disorder who were prescribed an antidepressant for at least 30 days in the year before pregnancy; 189 infants (8 per cent) were born with a major malformation. Paroxetine (36 per cent, sertraline (15 per cent) and venlafaxine (13 per cent) were the most frequently prescribed antidepressants in the first trimester. Compared with controls, antidepressant use during the first trimester was not associated with a higher risk of malformations and longer duration of treatment (more than 60 days) did not increase the risk.

A second observational study from Canada examined the effect of timing and duration of treatment with SSRIs during pregnancy (Br J Psychiatry 2008;192:383–43). A total of 1575 women treated during early (<185 days) pregnancy and 1925 treated in late pregnancy were identified by record linkage.

After adjusting for duration of exposure and maternal illness, outcomes did not differ significantly between early and late exposure. However, increased duration of exposure (after adjusting for maternal illness severity and regardless of the timing of exposure) was associated with increased risks of reduced gestational age, lower birth weight, low weight for gestational age and respiratory distress.

The third study, this time from the National Teratology Service in Newcastle upon Tyne, compared outcomes in 45 full‐term infants exposed to typical antipsychotics during pregnancy and 25 exposed to atypical antipsychotics with 38 non‐exposed infants (Br J Psychiatry 2008;192:333–7). Exposure to a typical antipsychotic was associated with significantly shorter median gestational age (273 vs 280 days) among boys and lower mean birth weight among girls. By contrast, atypical antipsychotics were associated with a higher incidence of children large for gestational age (20 per cent vs 2 per cent with typical agents and 3 per cent in controls).

The authors suggest that the risk of weight gain with atypicals may also apply to infants exposed in utero, but further study is needed.

Two pooled analyses of trials involving patients with moderate to severe Alzheimer's disease cover similar territory.

An analysis of six trials involved a total of 1826 patients (MMSE <20, mean 12) randomised to placebo or memantine 20mg per day (Int J Geriatr Psychiatry 2008;23:537–45); in two of these trials, memantine was added to established anticholinesterase therapy.

After 24 or 28 weeks, memantine was associated with superior total Neuropsychiatric Inventory score and significant improvements in individual scores for delusions, agitation/aggression and irritability/lability. Memantine also improved specific symptoms among patients who were symptomatic at baseline, though the absolute differences were modest, and reduced the emergence of symptoms among asymptomatic patients. The most consistent improvement occurred with agitation/aggression.

The second analysis (J Clin Psychiatry 2008;69:341–8) was of three trials included in the first. This subgroup had lower mean MMSE score (nine) and included one trial in which memantine was added to donepezil therapy. This analysis also found a significantly higher rate of improvement in a cluster of symptoms (agitation/ aggression, delusions and hallucinations) with memantine (58 vs 45 per cent with placebo at six months). This study showed a greater rate of decline among patients with behavioural disturbances compared with others with Alzheimer's disease and a lower rate of emergence of new behavioural symptoms with memantine compared with placebo (24 vs 37 per cent).

A study from Finland has compared memantine with escitalopram in the treatment of depression associated with alcohol dependence (J Clin Psychiatry 2008;69:392–9). Eighty people attending a treatment clinic were randomised to either drug (at a dosage of 20mg daily); just over 40 per cent were currently drinking and 40 per cent reported abstinence for one to three months. The rest had been abstinent for up to one year. About three‐quarters of participants completed the trial; after six months, baseline scores of anxiety and depression decreased in both groups with no significant differences between them. Quality of life outcomes improved equally in both groups, and cognitive function scores were unchanged.

Drugs that have been evaluated as treatments for pathological gambling include the SSRIs, opiate antagonists and lithium. Citing possible links between this disorder and bipolar disorder and dopaminergic dysfunction, US psychiatrists now report a trial of olanzapine (J Clin Psychiatry 2008;69:433–40).

They randomised 42 people meeting DSM‐IV criteria for pathological gambling to placebo or olanzapine 2.5–15mg daily (mean final dose 5mg daily). Over half of those assigned to olanzapine and over a quarter taking placebo did not complete the trial. After 12 weeks there was no difference in scores of gambling behaviour or response rates among the remaining 25 participants.

The GMC says doctors should take into account patients' beliefs when providing information about treatment, so that patients can make decisions that are truly informed. This is relevant to prescribing medicines that contain animal products such as beef or pork gelatine or stearic acid. A survey of psychiatrists' in parts of Yorkshire with diverse ethnic communities suggests that many take little account of faith when considering treatment (Psychiatric Bull 2008;32:179–82).

Only 40 per cent of the 95 surveys posted were returned. Most respondents were from non‐white ethnic groups (68 per cent) and trained overseas (58 per cent); over a third were consultants. Although 71 per cent said they were aware that some medicines contained animal products only 53 per cent believed it was necessary to discuss this with patients. About 80 per cent did not know that selected psychotropic drugs contain animal products and did not inform patients of the possibility. Encouragingly, 75 per cent who had not discussed this information in the past said they would now do so, though half said they would wait for the patient to raise the question – many believing that compliance would be adversely affected.

It can be difficult to improve the quality of prescribing and to sustain any gains over time but, say researchers in the Republic of Ireland, repeated intervention makes a difference (Psychiatric Bull 2008;32:183–6).

They identified six indicators (polypharmacy, use of thioridazine, high‐dose antipsychotics, maintenance use of a benzodiazepine or hypnotic, and routine use of anticholinergic agents) and assigned a score of one point to each indicator identified per patient. The scores were audited to produce an assessment of prescribing practice quality (PPQ).

The authors previously reported that providing a programme of education, clear guidelines and non‐pharmacological therapies had successfully reduced PPQ scores after one and two years. Five years later, they report that a continued emphasis on prescribing protocols and education during each new doctor's induction, combined with annual audits of prescribing, has sustained the early improvements.

The proportion of patients with at least one indicator was 27 per cent in 2006, down from 38 per cent in 2002 and 46 per cent in 2001; the mean PPQ scores were 0.33, 0.52 and 0.75 in 2006, 2002 and 2001 respectively. Similar improvements were achieved in prescribing for 163 long‐term service users who were included in all three audits. However, two indicators were refractory to change after an encouraging response in the first year: use of anticholinergics and benzodiazepines.

Investigations using biochemical markers have suggested that phenytoin may have a greater effect on bone mineral density (BMD) than other antiepileptic drugs (AEDs). A prospective study has now confirmed this is the case and raised concern about the safety of phenytoin in young women (Neurology 2008;70:1586–93).

Ninety‐three women (mean age 32 years) taking monotherapy with phenytoin, carbamazepine, lamotrigine or valproate were followed up for one year; exercise, pregnancy status and calcium and vitamin D intake were similar across treatment groups.

At baseline, BMD in the spine, femoral neck and total hip were similar. After one year, significant bone loss (2.6 per cent) occurred only among women taking phenytoin – but only at the femoral neck. These women had a pattern of biochemical markers suggestive of secondary hyperparathyroidism with increased bone turnover. Serum calcium concentrations were highest in women taking lamotrigine.

The authors comment that the rate of bone loss associated with phenytoin was eight times the normal rate for this age group. Women who begin long‐term treatment before the menopause may therefore be at increased risk of fractures.

A US study has revealed the astonishing number of medicines prescribed for patients with advanced dementia living in residential accommodation (Int J Geriatr Psychiatry 2008;23:490–6).

The mean number of medicines prescribed for 125 residents in three nursing homes was 14.6 (range 2–36) over a six‐month period. Indications included a wide range of co‐morbidities, the most frequently prescribed classes being antibiotics (88 per cent of residents), non‐opioid analgesics (83 per cent) and drugs for gastrointestinal disorders (82 per cent). Records for 88 residents who died showed an increase in prescribing opioids and a decrease in other analgesics, antibiotics, antipsychotics and dementia medications towards the time of death but the average number of prescribed medicines remained high.

Antipsychotics may be associated with early‐onset changes in glucose homeostasis in treatment‐naïve Asian men with schizophrenia (Acta Psychiatr Scand 2008;117: 342–7).

Psychiatrists in India randomised 99 patients (mean age 26 years) to treatment with olanzapine (mean dose 16.5mg daily), risperidone (4.4mg daily) or haloperidol (13.4mg daily). Two‐hour fasting blood glucose (FBG) and postprandial glucose (PPG) measurements were recorded at baseline and after treatment.

Baseline PPG, but not FPG, was elevated in men with schizophrenia compared with healthy controls, suggesting an underlying abnormality of glucose homeostasis. After six weeks, body weight and both FPG and PPG were increased. Comparing the antipsychotics, PPG at endpoint was lowest with haloperidol and greatest with olanzapine. FPG measurements were similar in the three drug groups.

There was also a significant increase in the incidence of diabetes (WHO definition) of 10 per cent in patients taking antipsychotics.

There is little information on the relative risks and benefits of different forms of drug treatment for opioid dependence. Investigators in Australia have now reported a retrospective record linkage study comparing methadone maintenance treatment with a naltrexone implant in heroin‐dependent subjects, using hospital morbidity and mortality data to assess outcomes over 3.5 years (Arch Gen Psychiatry 2008;65:457–65).

The cohort included 314 recipients of naltrexone and 522 methadone users. There were 18 deaths; six were drug‐related, including five among methadone users. Compared with the six months before treatment, naltrexone one was associated with a 77 per cent lower risk of opioid overdose and a 36 per cent lower risk of non‐overdose opioid morbidity after 3.5 years. By contrast, there was no significant change in opioid‐related events or admissions with methadone.

Both treatments were associated with a significantly increased risk of non‐opioid overdose after six months in older patients – in the case of naltrexone, a 16‐fold increase – but this did not persist at 3.5 years. Naltrexone, but not methadone, was also associated with increases in non‐opioid, non‐overdose morbidity at 3.5 years.

The authors suggest that methadone maintenance was associated with continuing opioid use and naltrexone with abuse of non‐opioids. The different risks and benefits should be balanced for each individual, they conclude.

GSK has announced a patient support service that aims to provide information and support to people with Parkinson's disease who are to be switched to the new prolonged‐release formulation of ropinirole – ReQuip XL.

The new service is optional and designed to complement NHS care and is to be offered via telephone by health professionals from Intrahealth based in Newcastle, which is contracted to provide the service.

Problems have arisen with the prescribing of the starter pack of the immediate‐release formulation of ropinirole. In some cases patients have received repeat prescriptions of the starter pack instead of the folpatients have not been titrated correctly up to a therapeutic dose. It is hoped the new service may help avoid the problem.

Prolonged‐release ropinirole is approved for the treatment of idiopathic Parkinson's disease (monotherapy and adjunct therapy) in patients already taking ropinirole immediate‐release tablets and in whom adequate symptomatic control has been achieved. Detailed advice on switching from immediate‐release to the prolonged‐release formulation of the drug is given in the Summary of Product Characteristics. Generally, the dose of ropinirole prolonged‐release tablets should be based on the total daily dose of immediate‐release formulation that the patient was receiving.

In a 24‐week double‐blind placebo‐controlled parallel group study of advanced stage Parkinson's disease patients, ropinirole prolonged‐release produced a significant reduction of 2.1 hours per day in awake time spent ‘off’ with no increase in troublesome dyskinesia during ‘on’ time compared with placebo when used as adjunct therapy to levodopa. Professor Carl Clarke, consultant neurologist at City Hospital, Birmingham, commented at a Press briefing to announce the launch of the new formulation and the patient support service that the additional time ‘on’ seen with the new formulation was probably a little more than would be expected with other adjunctive therapies.

GSK says ropinirole prolonged‐release tablets have been studied in 620 patients in controlled studies lasting over six months. And over 1.3 million patient years' experience exists for ropinirole immediate‐release three times daily in clinical practice. Copyright © 2008 Wiley Interface Ltd