Meta-analysis for trend estimation.

Grouped dose measures, heterogeneity and publication bias are three major problems for meta-analysis in trend estimation. In this paper, we propose a model that allows for arbitrarily aggregated dose levels, and show that the resulting estimates and standard errors can be quite different from those given by the usual assigned value method. Based on fitting a model to the funnel plot, we discuss a method for random-effects sensitivity analysis that deals with the problems of heterogeneity and publication bias. A meta-analysis of epidemiological studies on the effect of alcohol on the risk of breast cancer is used to illustrate the method. Our analysis suggests that the rate of increase in risk with alcohol consumption is substantially less than has been previously suggested.

A sensitivity analysis for publication bias in systematic reviews.

There is no simple method of correcting for publication bias in systematic reviews. We suggest a sensitivity analysis in which different patterns of selection bias can be tested against the fit to the funnel plot. Publication bias leads to lower values, and greater uncertainty, in treatment effect estimates. Two examples are discussed. An appendix lists the S-plus code needed for carrying out the analysis.

Reanalysis of epidemiological evidence on lung cancer and passive smoking.

OBJECTIVE: To assess the epidemiological evidence for an increase in the risk of lung cancer resulting from exposure to environmental tobacco smoke. DESIGN: Reanalysis of 37 published epidemiological studies previously included in a meta-analysis allowing for the possibility of publication bias. MAIN OUTCOME MEASURE: Relative risk of lung cancer among female lifelong non-smokers, according to whether her partner was a current smoker or a lifelong non-smoker. RESULTS: If it is assumed that all studies that have ever been carried out are included, or that those selected for review are truly representative of all such studies, then the estimated excess risk of lung cancer is 24%, as previously reported (95% confidence interval 13% to 36%, P<0.001). However, a significant correlation between study outcome and study size suggests the presence of publication bias. Adjustment for such bias implies that the risk has been overestimated. For example, if only 60% of studies have been included, the estimate of excess risk falls from 24% to 15%. CONCLUSION: A modest degree of publication bias leads to a substantial reduction in the relative risk and to a weaker level of significance, suggesting that the published estimate of the increased risk of lung cancer associated with environmental tobacco smoke needs to be interpreted with caution.

Meta-analysis, funnel plots and sensitivity analysis.

Publication bias is a major problem, perhaps the major problem, in meta-analysis (or systematic reviews). Small studies are more likely to be published if their results are 'significant' than if their results are negative or inconclusive, and so the studies available for review are biased in favour of those with positive outcomes. Correcting for this bias is not possible without making untestable assumptions. In this paper, a sensitivity analysis is suggested which is based on fitting a model to the funnel plot. Some examples are discussed.

Using regression models for prediction: shrinkage and regression to the mean.

The use of a fitted regression model in predicting future cases, either as a diagnostic tool or as an instrument for risk assessment is discussed. The regression to the mean effect implies that the future values of the response variable tend to be closer to the overall mean than might be expected from the predicted values. The extent of this shrinkage is studied for multiple and logistic regression models, and is found to be related to simple goodness-of-fit statistics of the original regression. Shrinkage is a particularly serious problem if the sample size is small and/or the number of covariates is large. Shrinkage of predictors is illustrated by two examples. A more general formulation is suggested.

Using generalized linear models (GLMs) to model errors in motor performance.

Because of differences in design factors, experiments in human motor performance sometimes produce a wide range in variability or consistency in a subject's individual errors. These differences in variation often lead to heterogeneity in the variance-covariance matrices between group factors, which prohibits the use of repeated-measures (RM) ANOVA or MANOVA techniques to analyze the error data. Provided certain conditions are met, however, each subject's individual errors can be collapsed into the summary error measures, constant error (CE) and variable error (VE), which can still provide a more than adequate description of the subjects' performance. This article proposes the appropriate conditions and the corresponding generalized linear models (GLMs) with which a subject's individual errors, recorded in short-term motor memory research, can be combined into the summary measures, CE and VE, which can be analyzed subsequently as the dependent variables in the experimental design. The CE scores can be modeled using GLMs without requiring the assumption of homogeneity of variances. Similarly, the VE scores can be modeled as a GLM, using a log-linear regression model that assumes a gamma distribution, rather than using a traditional analysis of variance (ANOV A) model that assumes an inappropriate normal error distribution for these scores. An example reveals that the analysis of VE scores, unlike the analysis of CE scores, is able to differentiate between group practice methods. These differences tend to be underestimated by traditional ANOVA methods, however. Differences between the ANOV A and GLM analyses of the VE scores are further clarified by simulation. Based on differences like those observed in the example, when simulated VE scores were analyzed, assuming both a normal and a gamma distribution, the power of the gamma tests was found to be superior to the normal analyses in all but a small range of cases, in which such differences were found to be negligible. Hence, it is only by declaring the VE scores to have a GLM with a gamma distribution that the anticipated group practice differences can be properly identified.

Schizophrenia, bipolar disorder and depression. A discriminant analysis, using 'lifetime' psychopathology ratings.

Discriminant and canonical variate analyses were performed using 302 patients, on whom ratings of lifetime psychopathology and course of illness has been made. DSM-III diagnoses were used to form the criterion groups. Bipolar disorder emerged as a distinct grouping, but there are reasons for dissatisfaction with its definition. The remaining patients formed a 'schizodepressive continuum', but this also had a tendency to bimodality. It is possible that the distinction between schizophrenia and depression was obscured by inadequacies in the data and the inclusion of excessive numbers of patients with schizoaffective depression in this study.

Community care: does it reduce the need for psychiatric beds? A comparison of two different styles of service in three hospitals.

The admission pattern of under-65s during the first seven years of Kidderminster's District General Hospital Psychiatric Department was compared with the last seven years of the mental hospital (Powick Hospital) which it superseded in July 1978, and with its nearest mental hospital (Barnsley Hall Hospital). In Kidderminster the average length of patient's hospital stay was half that at Barnsley Hall. There were more admissions per head of population, more which lasted less than a week, and more patients with over six admissions to Kidderminster than Barnsley Hall. The retrospective comparison with Powick Hospital showed that these differences were not due to the superior community facilities in Kidderminster since there was no real change in hospital bed use with the development of community services; this supports the view that there is a bed-rock of illness which will always need in-patient care however comprehensive the community resources.

Record linkage: statistical models for matching computer records.

"We wish to measure the evidence that a pair of records relates to the same, rather than different, individuals. The paper emphasizes statistical models which can be fitted to a file of record pairs known to be correctly matched, and then used to estimate likelihood ratios. A number of models are developed and applied to U.K. immigration statistics. The combination of likelihood ratios for possibly correlated record fields is discussed." A series of comments on the paper is also included, as well as a reply to those comments by the author (pp. 312-20).

The Facial Appearance Sorting Test (FAST): an aid to the selection of patients for rhinoplasty.

A simple test which requires a few minutes to administer and score is described as an aid to the selection of patients for rhinoplasty. The sorting of cards evaluating appearance is performed consistently over a period of time by individual subjects and by different groups of subjects. A numerical formula is proposed which allows for deviation from the common system of evaluating appearance in measuring support for an application for rhinoplasty.

Reshaping the psyche. The concurrent improvement in appearance and mental state after rhinoplasty.

Rhinoplasty patients and matched elective-surgery controls completed the Facial Appearance Sorting Test, the General Health Questionnaire, a Repertory Grid and the Masculinity/Femininity Scale. Rhinoplasty applicants perceived appearance similarly to, and downrated their own appearance to the same extent as, controls. Impaired appearance and psychiatric symptoms are integral parts of the 'rhinoplasty applicant syndrome', but the degree to which they occur is not positively correlated. Interviews and tests were repeated 6 months after operation, when marked improvement in appearance was reported by the rhinoplasty patients, associated with the reduction of psychiatric-symptom scores. Control subjects showed no change.

Electrophysiological and hormonal responses to three types of electroconvulsive therapy.

Patients were treated with three different types of electroconvulsive therapy in random order in the first three sessions of their courses. Anaesthetic and muscle relaxants were administered in standard doses, and treatment applied a set time after bolus injection. Electroencephalographic recordings were undertaken before, during, and after treatment, and hormonal measures before and after. The responses to treatment differed between the three types of current in both EEG and prolactin measures. It is suggested that the quality of the convulsive response varies with different currents.

Total estriol in serum and plasma as determined by radioimmunoassay.

We present and discuss the results of a statistical study of a set of radioimmunoassay data obtained from parallel determinations of the total estriol content of samples of serum and plasma prepared from 98 specimens of whole blood, taken from pregnant woman. Estriol concentrations in the serum exceed by an average of 7% the values found for the plasma prepared from the same whole-blood specimen. The clinical significance of this analytical finding is discussed, and we conclude with the recommendation that serum is the fluid of choice for this assay.

The actions of 5-, 12-, and 15-HETE on tracheobronchial smooth muscle.

In view of the likelihood that monohydroxyeicosatetraenoic acids (HETEs) are produced in asthma we investigated the actions of these lipoxygenase products on a range of tracheobronchial smooth muscle preparations. 5- and 15-HETE contracted human isolated bronchial muscle and guinea-pig lung strip but were less potent and less effective than histamine. In contrast 5-, 12- and 15-HETE were inactive on guinea-pig and cat tracheae. In addition to its direct action, subthreshold doses of 5-HETE (0.001-0.01 micrograms ml-1) potentiated histamine-induced contractions of human isolated bronchial muscle. The latter effect may suggest that 5-HETE plays a significant role in the airway hyperreactivity commonly associated with asthma.

Problem areas and mental and behavioral status in schizophrenia, neurosis, and depression.

A 5- to 8-year follow-up of first admissions to a local psychiatric service enabled the comparison of patients with and without symptoms for the presence of current problem areas. Whereas first admission was associated with a temporary high level of reports of problems, the 6-month and 5- to 8-year levels were stable. Relative risk of onset following an event compared with spontaneous onset was likewise stable for the different diagnoses, significant only for depressive states and not for schizophrenia or neurosis, despite higher overall reporting of problems in the latter. The type of depression emerging after problems was similar to that occurring spontaneously.

2-decarboxy-2-hydroxymethyl prostaglandin E1 (TR4161), a prostaglandin bronchodilator of low tracheobronchial irritancy.

2-Decarboxy 2-hydroxymethyl prostaglandin E1 (TR4161) relaxed isolated quinea-pig trachea with about double and relaxed human isolated bronchial muscle with about one half the potency of PGE1. In conscious restrained cats an aerosol of TR4161 was about 100-1000 times less active than PGE1 in inducing tracheobronchial irritation. When given intravenously or by aerosol to the anaesthetised spontaneously breathing guinea-pig, TR4161 was approximately equipotent with PGE1 in inhibiting histamine-induced bronchoconstriction and in reducing basal inherent tone. The onset and duration of the bronchodilator effects of TR4161 administered intravenously, however, were significantly longer than those of PGE1. In conscious quinea-pigs, TR4161 by aerosol was approximately three times more potent than PGE1 in preventing histamine-induced convulsions, whereas only TR4161 was active in this test system when the test drugs were administered orally. These observations indicate that TR4161 might be therapeutically useful as a non-irritant prostaglandin bronchoidilator in conditions of airway obstruction.

Treatment settings in psychiatry: long-term family and social findings.

Treatment in a psychiatric unit, as distinct from a psychiatric hospital, held no benefits in the long term (5-8 years after admission) as far as the patient's mental and behavioural status and employment, or the family's burdens, health needs or attitude were concerned. Schizophrenic patients from both settings received more hospital treatment and medication, were more often unemployed, and had more adverse effect on relatives' health than neurotic or depressive patients. Hospital attendence identified relatives who expressed a need for support.

Tracheobronchial irritancy of inhaled prostaglandins in the conscious cat.

A novel test was developed to measure the tracheobronchial irritant activity of inhaled prostaglandins. Conscious restrained cats were challenged with separate aerosols of PGE1, PGF2alpha, acetylcholine or isoprenaline. All of the aerosols except isoprenaline caused coughing in a concentration related manner. Tolerance developed very quickly to the tracheobronchial irritation and lasted 1-2 days for PGE1 and less than 1 day for PGF2alpha and acetylcholine. When a 3 day interval between each aerosol challenge was used, PGF2alpha was approximately 700 times more potent than acetylcholine as a tracheobronchial irritant. The highest PGE1 aerosol concentration (500microgram/ml) also caused sedation, diarrhoea and salivation. This test probably provides a useful method for evaluating the tracheobronchial irritant activity of potential prostaglandin bronchodilator analogues and for investigating the mechanism of action of prostaglandin induced tracheobronchial irritancy.

Endogenous inhibitor of prostaglandin synthetase.

Mammalian serum and plasma contain an endogenous inhibitor of prostaglandin synthetase (EIPS). Human plasma fractions rich in EIPS show anti-inflammatory activity in vivo. In rats, glucocorticoids raise EIPS activity of plasma and serum. These findings suggest the existence of a natural mechanism of controlling prostaglandin synthesis, possibly related to corticosteroid action.