Meta-analysis

The communication of suicidal intentions: a meta-analysis

BACKGROUND

Among the myths that are often cited about suicide is that 'people who talk about killing themselves rarely die by suicide', but the evidence seems to contradict this statement. The aim of this study was to conduct a meta-analysis of studies reporting a prevalence of suicide communication (SC), and to examine the diagnostic accuracy of SC towards suicide in case-control reports.

METHOD

Eligible studies had to examine data relative to completed suicides and report the prevalence of SC. Data relative to sample characteristics, study definition, modality and recipient of the SC were coded.

RESULTS

We included 36 studies, conducted on a total of 14 601 completed suicides. The overall proportion of SC was 44.5% [95% confidence interval (CI) 35.4-53.8], with large heterogeneity (I 2 = 98.8%) and significant publication bias. The prevalence of SC was negatively associated with the detection of verbal communication as the sole means of SC and, positively, with study methodological quality. Based on seven case-control studies, SC was associated with an odds ratio of 4.66 for suicide (95% CI 3.00-7.25) and was characterized by sufficient diagnostic accuracy only if studies on adolescents were removed.

CONCLUSION

Available data suggest that SC occurs in nearly half of subjects who go on to die by suicide, but this figure is likely to be an underestimate given the operational definitions of SC. At present, SC seems associated with overall insufficient accuracy towards subsequent suicide, although further rigorous studies are warranted to draw definite conclusions on this issue.

Does Fibronectin Reduce Mortality in Sepsis? A Meta-Analysis

We examine the effect of fibronectin administration on mortality in sepsis. An English language literature search using MEDLINE, EMBASE, SCISEARCH, and compu terized databases of unpublished research in addition to bibliographic review was performed, and duplicate independent review of 41 articles by two observers identified six relevant randomized clinical trials. Study quality and descriptive information concerning the pop ulation, intervention, and outcome measurements were assessed. Three of six trials demonstrated favorable mor tality trends, but none reached statistical significance. Results of the 282 patients in all six trials yield a com mon odds ratio of 0.97 and a 95% confidence interval of 0.58 to 1.61. These results do not show a statistically significant effect favoring treatment, but neither do they exclude a clinically important benefit. The benefit of fibronectin in decreasing mortality in sepsis is not sup ported by these studies. However, methodological heterogeneity and the possibility of a Type II error, as well as our understanding of the role of fibronectin in the pathophysiology of sepsis, do not allow for a firm negative conclusion regarding the therapeutic efficacy of fibronectin in sepsis. A large, rigorously designed trial should be considered once investigators can reach a consensus regarding current methodological and patho physiological controversies.

WITHDRAWN: Steroids for treating tuberculous meningitis

BACKGROUND

Even though corticosteroids have been used alongside antituberculosis drugs for tuberculous meningitis (TBM) since the 1950s their role remains controversial. Some believe corticosteroids improve outcome while others point to the lack of supportive evidence. In patients who are immunocompromised because of HIV infection the risks and benefits of steroids are unknown.

OBJECTIVES

To assess the effects of steroids on death and disability in patients with TBM.

SEARCH STRATEGY

We searched the Cochrane Infectious Diseases Group specialized trials register (February 2005), The Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 1, 2005), MEDLINE (1966 to February 2005), EMBASE (1980 to February 2005), and LILACS (February 2005).

SELECTION CRITERIA

Randomised controlled trials of steroids in people on TB treatment for TBM.

DATA COLLECTION AND ANALYSIS

Two independent reviewers applied study selection criteria, assessed methodological quality and extracted data.

MAIN RESULTS

Six trials of 595 patients met the inclusion criteria. No study described allocation concealment. Steroids were associated with fewer deaths (relative risk [RR] 0.79; 95% confidence interval [CI] 0.65 to 0.97) and a reduced incidence of death and severe residual disability (RR 0.58, 95% CI 0.38 to 0.88). Subgroup analysis suggests an effect on mortality in children (RR 0.77, 95% CI 0.62 to 0.96) but the results in a smaller number of adults are inconclusive (RR 0.96, 95% CI 0.50 to 1.84). There is little evidence that the severity of disease influences the effects of steroids on mortality.

AUTHORS' CONCLUSIONS

Adjunctive steroids might be of benefit in patients with TBM. However, existing studies are small, and poor allocation concealment and publication bias may account for the positive results found in this review. No data are available on the use of steroids in HIV positive persons. Future placebo-controlled studies should include patients with HIV infection and should be large enough to assess both mortality and disability.

Death from suicide versus mortality from epilepsy in the epilepsies: a meta-analysis

OBJECTIVE

Epilepsy is associated with an increased risk of mortality, which, however, is rarely due to the epilepsy itself; suicide, on the other hand, is a chief cause of death among persons with epilepsy. We conducted a meta-analysis to compare data reported in representative studies of suicide mortality in epilepsy with data on mortality from epilepsy in the general population.

METHODS

We searched Index Medicus to 2006 through MedLine. We also searched the World Health Statistics Annual to ascertain rates of mortality from epilepsy in the age groups indicated in the studies on patients with epilepsy for specific years and countries.

RESULTS

We selected 30 studies comprising 51,216 persons, 188 of whom committed suicide and died. Results obtained for each study were processed together to calculate, per 100,000 individuals in the general population per year, the mean number of suicide deaths expected in persons with epilepsy with respect to the current prevalence of epilepsy. We found that the number of suicide deaths among persons with epilepsy is the same as the number of deaths from epilepsy, suggesting that the former are not included in mortality rates for epilepsy. Also, we found that, according to data derived from cohorts we selected, 32.5% of all deaths of persons with epilepsy are due to suicide and at least 13.5% of all registered suicides are committed by these persons. Study findings may not generalize to other samples, settings, and treatments, thus perhaps tending to exaggerate the phenomenon actually attainable under broader clinical conditions.

CONCLUSIONS

Our meta-analysis indicates that suicide deaths in persons with epilepsy are a disturbingly frequent phenomenon that should be addressed to reduce mortality among patients with epilepsy dramatically. In particular, mortality rates for persons with epilepsy do not include mortality from suicide, greatly underestimating death rates and the need for suicide prevention strategies for these patients.

Dose-response relationship of duloxetine in placebo-controlled clinical trials in patients with major depressive disorder

RATIONALE

The optimal dose for achieving the maximum antidepressive effect of selective serotonin reuptake inhibitors (SSRIs) or serotonin-noradrenalin reuptake inhibitors (SNRIs) remains a controversial issue. The varying sensitivity of scales that measure the severity of depression is one of the many factors affecting the evaluation of the dose-response relationship with antidepressants.

OBJECTIVES

To determine if the 6-item Hamilton rating scale for depression (HAM-D6) demonstrates a clearer association between dose and antidepressive effect compared with the 17-item Hamilton rating scale for depression (HAM-D17) for fixed doses of duloxetine hydrochloride (40, 60, 80, and 120 mg daily) from six double-blind, randomized, placebo-controlled clinical trials assessing safety and efficacy in the acute treatment of patients with DSM-IV-defined major depressive disorder (MDD).

METHODS

Mantel-Haenszel adjusted effect sizes were determined by dose for change from baseline to endpoint in HAM-D6 and HAM-D17 scores from the six studies. To confirm, assessments were repeated on the subset of the population corresponding to the 70% of patients with the longest duration of treatment regardless of study, treatment, dose, geography, or completion status.

RESULTS

For the majority of assessments, HAM-D6 effect sizes were numerically larger than those estimated from the HAM-D17. Findings support that duloxetine 60 mg daily is the best effective dose.

CONCLUSIONS

In this assessment of patients with MDD, the HAM-D6 was shown to be more sensitive compared with the HAM-D17 at detecting treatment effects. These findings are consistent with published results of other effective antidepressants.

Suicide after surgical treatment in patients with epilepsy: a meta-analytic investigation

Suicide is a major issue in surgically treated epileptic patients. A meta-analysis was performed comparing suicides in a sample of such patients and in the general population. The Index Medicus and the World Health Statistics Annual were searched to ascertain the suicide rates in the age groups indicated in the studies of epileptic patients for specific years and country. 11 studies were selected, comprising 2,425 patients, 24 of whom committed suicide. Data obtained for each study were processed together to calculate the mean number of suicides per 100,000 individuals with surgically treated epilepsy for each year. This meta-analysis shows that suicide in patients with epilepsy after surgical treatment is more frequent than in the general population. Results are discussed with particular attention to possible causative factors.

Suicide in the epilepsies: a meta-analytic investigation of 29 cohorts

OBJECTIVES

Suicide in epilepsy is a major cause of death and is very often underestimated. We conducted a meta-analysis to compare data reported in representative studies of suicide in epilepsy with data for the general population.

METHODS

We searched the Index Medicus until 2005 through MedLine. We also searched the World Health Statistics Annual to ascertain the suicide rates in the age groups indicated in the studies on epileptic patients for specific years and country.

RESULTS

We selected 29 studies comprising 50,814 patients, 187 of whom committed suicide. Results obtained for each study were processed together to calculate the mean number of suicides per 100,000 individuals suffering from epilepsy for each year.

CONCLUSIONS

Our meta-analysis shows that suicide in patients with epilepsy is more frequent than in the general population. Nevertheless, a number of cohorts of epileptic patients had a suicide rate lower than that of the general population. Possible explanations for this heterogeneous trend and risk factors are discussed.

Suicide in borderline personality disorder: a meta-analysis

Suicide is the major cause of death among patients with borderline personality disorder; however, the literature on completed suicides in such disorder is inconclusive, as suicide rates vary greatly among cohorts of patients. We searched MedLine, Excerpta Medica and PsycLit from 1980 to 2005 to identify papers dealing with suicide in borderline personality disorder. We also searched the World Health Statistics Annual to ascertain the suicide rate in the age groups for specific years and country. We selected eight studies comprising 1179 patients with a diagnosis of borderline personality disorder. Of these patients, 94 committed suicide. Results obtained for each study were processed together to calculate the mean figure for each year of suicides for 100,000 individuals suffering from borderline personality disorder. Our meta-analysis shows that suicide among patients with borderline personality disorder is more frequent when compared with the general population. All study analyses reported that patients with borderline personality disorder committed suicide more often than their counterparts in the general population. Suicide seems more alarming in the first phases of follow-up than during chronic phases of illness.

Gender influence on the dose-ranging of a low-dose perindopril-indapamide combination in hypertension: effect on systolic and pulse pressure

BACKGROUND

Dose-ranging of antihypertensive agents have been done to optimize diastolic blood pressure (DBP) reduction, but with little information on systolic (SBP), mean (MBP), or pulse (PP) pressures. A low-dose combination of perindopril (Per) and indapamide (Ind) has been shown to reduce more SBP than atenolol for the same DBP reduction. However, the possible influence of gender on this finding has never been tested.

PURPOSE

A database of five randomized, double-blind, dose-ranging studies was established to determine the optimal dose of the Per/Ind combination in hypertensive men and women. A total of 2907 patients were treated by either placebo or various combinations associating Per (2, 4, 8 mg) and Ind (0.625, 1.25, 2.5 mg).

RESULTS

In the overall population, there was a significant dose-response relationship (P < 0.001) for doubling the dose of Per 2/Ind 0.625 mg up to Per 8/Ind 2.5 mg with a progressive fall in SBP, DBP, MBP. When men and women were analyzed by dose, SBP, DBP and MBP (but not PP) decreased significantly more in women than in men until the Per 4/Ind 1.25 dosage was reached. Thereafter, with higher dosages, generating a slight but significant hypokalemia, the finding was reversed, resulting in a gender interaction in the overall population.

CONCLUSION

In hypertensive subjects, the low-dose combinations Per 2/Ind 0.625 and Per 4/Ind 1.25 are the most effective in reducing blood pressure and avoiding hypokalemia. This effect is more pronounced in women, in which increased SBP and PP are predominant hemodynamic features.

A comparison of the efficacy and tolerability of adapalene 0.1% gel versus tretinoin 0.025% gel in patients with acne vulgaris: a meta-analysis of five randomized trials

The purpose of this meta-analysis was to determine if adapalene 0.1% gel (Differin) provided superior efficacy and better tolerability than tretinoin 0.025% gel in the treatment of acne vulgaris. All comparative studies, both published and unpublished, from the United States and Europe, that fulfilled rigorous protocol criteria (multicentre, randomized, investigator-blind) were used. Five comparative studies met these criteria. In total, the meta-analysis evaluated 900 patients (450 treated with adapalene 0.1% gel, 450 treated with tretinoin 0.025% gel) with mild-to-moderate acne from the combined clinical trials. To avoid study bias, the meta-analysis used an intention-to-treat analysis. Statistical methodology for the meta-analysis included analysis of covariance, analysis of variance and Cochran-Mantel-Haenszel test. All statistical tests were two-sided, with the 0.05 probability level used to establish statistical significance, and 95% confidence intervals used to assess equivalence. Adapalene demonstrated equivalent efficacy to tretinoin in terms of reducing total lesion count. Adapalene demonstrated more rapid efficacy, as evidenced by a significant difference in the reduction of inflammatory and total lesions at week 1. Adapalene also demonstrated considerably greater local tolerability at all evaluation periods. The findings from this meta-analysis suggest that adapalene 0.1% gel constitutes a pharmacologic advance over such classic retinoids as tretinoin for the treatment of acne vulgaris.

Efficacy of screening mammography among women aged 40 to 49 years and 50 to 69 years: comparison of relative and absolute benefit

In randomized controlled trials, screening mammography has been shown to reduce mortality from breast cancer about 25% to 30% among women aged 50 to 69 years after only five to six years from the initiation of screening. Among women aged 40 to 49 years, trials have reported no reduction in breast cancer mortality after seven to nine years from the initiation of screening; after 10 to 14 years there is a 16% reduction in breast cancer mortality. Given that the incidence of breast cancer for women aged 40 to 49 years is lower and the potential benefit from mammography screening smaller and delayed, the absolute number of deaths prevented by screening women aged 40 to 49 years is much less than in screening women aged 50 to 69 years. Because the absolute benefit of screening women aged 40 to 49 years is small and there is concern that the harms are substantial, the focus should be to help these women make informed decisions about screening mammography by educating them of their true risk of breast cancer and the potential benefits and risks of screening.

Pharmacological treatment of severely depressed patients: a meta-analysis comparing efficacy of mirtazapine and amitriptyline

Efficacy data were available from 405 severely depressed patients (baseline 17-item Hamilton Rating Scale for Depression-HAMD scores > or = 25) participating in randomized, double-blind, amitriptyline-controlled studies of mirtazapine. Main efficacy variable were changes from baseline in the group mean 17-item HAMD scores and responder rates. Secondary efficacy variables were changes in depressed mood item on the HAMD and in factors derived from the 17-item HAMD scale. Treatment with either mirtazapine or amitriptyline resulted in robust reductions of baseline HAMD scores and in similar and high percentages of responders. Both drugs produced favourable effects on depressed mood and on symptoms commonly associated with depression, such as anxiety, sleep and vegetative disturbances. There were neither statistically significant nor clinically relevant differences between mirtazapine and amitriptyline at any assessment point nor at endpoint. The results demonstrate that the new antidepressant mirtazapine and the tricyclic antidepressant amitriptyline are equally effective in the treatment of severely depressed patients.

Meta-analysis and investigation of anti-infective therapies

Meta-analysis offers an alternative approach to examining specific sub-groups, such as classes of antibiotics, types of bacteria as well as patient groups whose response to the agent under investigation may not be equivalent. In this review, the strengths and pitfalls of meta-analyses of anti-infective therapies will be discussed.

Meta-analysis of randomized, double-blind, placebo-controlled, efficacy and safety studies of mirtazapine versus amitriptyline in major depression

A meta-analysis was performed on efficacy and safety data from 4 randomized, double-blind, 6-week, single-center studies comparing mirtazapine (n = 194; 5-35 mg/day) with amitriptyline (n = 193, 40-280 mg/day) and placebo (n = 193) in outpatients with a DSM-III diagnosis of major depressive episode. On all the main efficacy variables both active drugs consistently produced significantly greater improvements and significantly greater percentages of responders or remitters than placebo. The meta-analysis of adverse events shows that mirtazapine was better tolerated than amitriptyline, particularly with respect to anticholinergic and cardiac adverse events. There were no differences between mirtazapine and placebo regarding the incidence of serotonergic adverse events. In conclusion, the results of this meta-analysis demonstrate that mirtazapine is as effective as amitriptyline but has a better tolerability profile.

Meta-analysis and prospective meta-analysis in childhood leukemia clinical research

In this paper, we consider the role of meta-analysis and 'prospective meta-analysis' studies in childhood acute lymphocytic leukemia (ALL). In this issue, Valsecchi and Masera [1] give a thoughtful discourse, generally favorable to this approach. This article presents the opposite point of view. The aims of our article are to present the implications in clinical, rather than biostatistical terms, and to provide an extensive literature review of the subject of meta-analysis. We conclude that treatment assessments, resulting from meta-analysis of closed studies (retrospective) should be met with healthy skepticism. Trials requiring international resources should be true intergroup trials with a single coordinating center, rather than prospective meta-analysis, unless it is a question grafted onto each group's own research agenda. For example, each group might ask its own systemic control question, but a CNS protection question is asked collectively.

[Inclusion of research quality in meta-analyses]

The objective of this review is how study quality is taken into account in the synthesis of information from several studies. Meta-analysis has been criticized for failing into account differential quality of evidence in primary studies, allowing poor studies to drive out good science by weight of numbers. Notwithstanding there is presently no agreement on the inclusion of study quality score in quantitative meta-analysis; on the other hand, agreement does exist on knowing study quality. However, and although there is a certain level of consensus about what a good study is and on what characteristics it should have, no evaluation protocol on study quality is uniformly accepted. Furthermore, there are no data on the reliability and validity of quality questionnaires. There are several procedures to include study quality in meta-analysis: a) it is generally accepted that bias-free studies should be pooled only; b) most researches reject to include quality scores in the weight; and c) no agreement is reached on the use the quality scores as a covariate in graphical methods and regression analysis. It may be superfluous or misleading.

Critical evaluation of clinical research

This focused review contains a suggested core of material that will help residents or practicing physiatrists critically review research papers published in the medical literature. Before accepting the results of a clinical trial, physiatrists must critique the experimental methods and study design carefully to decide whether to include these new ideas into their clinical practice. Medical research relies on statistical methodology, and statistics pervade the medical literature. This article begins with an introduction to rudimentary statistics. Fortunately, most studies depend on a rather small body of statistical concepts. The elements of experimental design--clinical trials, randomization, single-subject design, meta-analysis, epidemiological studies--are presented in a concise review. Finally, the elements of statistics and experimental design are integrated into a step-by-step method strategy for reading the medical literature.

Fluoride varnishes (Duraphat): a meta-analysis

In the present article we analyze a series of studies designed to detect the caries preventive effect of Duraphat by means of meta-analysis. Meta-analysis is a collection of statistical methods designed to investigate and to summarize a series of investigations. It may be a valuable tool to complement traditional narrative reviews. During the last years meta-analysis has attracted increasing interest in sociology, psychology and medicine. In order to find previous studies concerned with the clinical effects of Duraphat we applied a systematic literature search. Papers were included independent of results when they fulfilled a checklist of well defined methodological selection criteria. In order to aggregate the results of the Duraphat-studies we used different complementary statistical approaches: Firstly, the so-called file drawer problem is considered. This may help to get a better insight into the problem of underreporting non-significant results or publication bias. It was found to be very unlikely that underreporting of non-significant results could reverse the conclusion into an overall null-result. After that, the inhomogeneity between studies is investigated. The overall variation of caries reduction R is separated into two components: A between study component of variance and a variance pertaining to the individual studies (random effects model). It was found that the overall variation is dominated by the between studies variation and not by the sampling variation. Due to the pronounced variation between studies the confidence interval of the overall effect size (R = 0.38) is quite large (95%-Cl: 0.19-0.57).(ABSTRACT TRUNCATED AT 250 WORDS)

Quality judgements by patients on general practice care: a literature analysis

Patient report seems to be a workable method for quality assurance. For the purpose of the development of such a method, a review was made of 40 studies into patient judgements on general practice care. Apparently, many aspects of care are only rarely included in patient report and the patients themselves are hardly ever involved in selecting these aspects. Furthermore, there appears to be a large variety in the methods that are used. The conclusion must be that in the field of general practice care only little progress has been made in the development of patient report as a method for quality assurance.

[Meta-analysis in epidemiology (1): the general characteristics]

In the present work the general characteristics, aims and applications of meta-analysis in public health are described. The general rules and the relevant steps for the development of this type of studies are considered. These points are illustrated with a number of real examples and the advantages and limitations of the different methodological options are commented.

The mild hypertension enigma

This study examines the problem of the correlation between mild arterial hypertension and cardiovascular damage. The authors examine the results of the most important trials carried out and, on the basis of their evaluations, suggest the need to review the current clinical policy of considering mild arterial hypertension as an important risk factor directly related to cardiovascular disease. Since the therapeutic trials carried out on mild hypertension did not substantially reduce the total and cardiac mortality rate, it seems to be probable that arterial hypertension is a progression acceleration marker of atheromatous disease. According to this theory, a therapy which aims merely at returning the pressure values to normal limits will probably not change the natural course of the atheromatous process.

An overview of statistical issues and methods of meta-analysis

A meta-analysis is a statistical analysis of the data from some collection of studies in order to synthesize the results. In this paper we discuss issues that frequently arise in meta-analysis and give an overview of the methods used, with particular attention to the use of fixed- and random-effects approaches. The methods are then applied to two sample datasets.

Meta-analysis of randomized, controlled clinical trials of antibiotic prophylaxis in biliary tract surgery

In this study all available clinical trials of antibiotic prophylaxis in biliary tract surgery, published from 1965 to 1988, were examined. Results of 42 randomized, controlled trials (4129 patients), in which a group of patients treated with antibiotics was compared with a group of patients not treated with antibiotics, were pooled. Wound infection rates in the control groups range from 3 to 47 per cent and are 15 per cent overall. The overall difference in infection rates is 9 per cent in favour of antibiotic treatment (95 per cent confidence interval 7-11 per cent), while the common odds ratio is 0.30 (95 per cent confidence interval 0.23-0.38). Subgroup meta-analysis showed a significant stronger protective effect in high risk patients, while the timing of wound inspection (i.e. early in hospital or late at follow-up) markedly influenced the treatment effect reported. Comparison of wound infection rates in patients treated with first generation versus second or third generation cephalosporins (11 trials, 1128 patients), as well as single-dose versus multiple-dose regimens (15 trials, 1226 patients) did not reveal any significant effect (P greater than 0.05) in each trial separately as well as in the overall comparison. The results indicate that there is evidence against further use of no-treatment controls and that the choice of treatment regimen can largely be made on the basis of cost.