Interpreting frequentist hypothesis tests: insights from Bayesian inference.

Randomized controlled trials are one of the best ways of quantifying the effectiveness of medical interventions. Therefore, when the authors of a randomized superiority trial report that differences in the primary outcome between the intervention group and the control group are "significant" (i.e., P ≤ 0.05), we might assume that the intervention has an effect on the outcome. Similarly, when differences between the groups are "not significant," we might assume that the intervention does not have an effect on the outcome. Nevertheless, both assumptions are frequently incorrect.In this article, we explore the relationship that exists between real treatment effects and declarations of statistical significance based on P values and confidence intervals. We explain why, in some circumstances, the chance an intervention is ineffective when P ≤ 0.05 exceeds 25% and the chance an intervention is effective when P > 0.05 exceeds 50%.Over the last decade, there has been increasing interest in Bayesian methods as an alternative to frequentist hypothesis testing. We provide a robust but nontechnical introduction to Bayesian inference and explain why a Bayesian posterior distribution overcomes many of the problems associated with frequentist hypothesis testing.Notwithstanding the current interest in Bayesian methods, frequentist hypothesis testing remains the default method for statistical inference in medical research. Therefore, we propose an interim solution to the "significance problem" based on simplified Bayesian metrics (e.g., Bayes factor, false positive risk) that can be reported along with traditional P values and confidence intervals. We calculate these metrics for four well-known multicentre trials. We provide links to online calculators so readers can easily estimate these metrics for published trials. In this way, we hope decisions on incorporating the results of randomized trials into clinical practice can be enhanced, minimizing the chance that useful treatments are discarded or that ineffective treatments are adopted.

RéSUMé: Les études randomisées contrôlées constituent l'un des meilleurs moyens de quantifier l'efficacité des interventions médicales. Par conséquent, lorsque les autrices et auteurs d'une étude randomisée superiorité signalent que les différences dans le critère d'évaluation principal entre le groupe d'intervention et le groupe témoin sont « significatives ¼ (c.-à-d. P ≤ 0,05), nous pourrions supposer que l'intervention a un effet sur le critère d'évaluation. De même, lorsque les différences entre les groupes ne sont « pas significatives ¼, nous pourrions supposer que l'intervention n'a pas d'effet sur le critère d'évaluation. Pourtant, ces deux hypothèses s'avèrent souvent incorrectes.Dans cet article, nous explorons la relation qui existe entre les effets réels d'un traitement et les déclarations de signification statistique fondées sur les valeurs P et les intervalles de confiance. Nous expliquons pourquoi, dans certaines circonstances, la probabilité qu'une intervention soit inefficace lorsque P ≤ 0,05 dépasse 25 % et la probabilité qu'une intervention soit efficace lorsque P > 0,05 dépasse 50 %.Au cours de la dernière décennie, nous avons assisté à un intérêt croissant pour les méthodes bayésiennes comme alternative aux tests d'hypothèses fréquentistes. Nous proposons une introduction robuste mais non technique à l'inférence bayésienne et expliquons pourquoi une distribution postérieure bayésienne surmonte bon nombre des problèmes associés aux tests d'hypothèses fréquentistes.Malgré l'intérêt actuel pour les méthodes bayésiennes, les tests d'hypothèses fréquentistes restent la méthode par défaut pour l'inférence statistique en recherche médicale. Par conséquent, nous proposons une solution provisoire au « problème de signification ¼ basée sur des mesures bayésiennes simplifiées (par exemple, facteur de Bayes, risque de faux positifs) qui peuvent être rapportées en même temps que les mesures traditionnelles des valeurs P et des intervalles de confiance. Nous calculons ces paramètres pour quatre études multicentriques bien connues. Nous fournissons des liens vers des calculatrices en ligne afin que les lectrices et lecteurs puissent facilement estimer ces mesures pour les études publiées. De cette façon, nous espérons que les décisions sur l'intégration des résultats des études randomisées dans la pratique clinique pourront être améliorées, minimisant ainsi le risque que des traitements utiles soient rejetés ou que des traitements inefficaces soient adoptés.

False-positive and false-negative risks for individual multicentre trials in critical care.

Background: In medical research, null hypothesis significance testing (NHST) is the dominant framework for statistical inference. NHST involves calculating P-values and confidence intervals to quantify the evidence against the null hypothesis of no effect. However, P-values and confidence intervals cannot tell us the probability that the hypothesis is true. In contrast, false-positive risk (FPR) and false-negative risk (FNR) are post-test probabilities concerning the truth of the hypothesis, that is to say, the probability a real effect exists. Methods: We calculated the FPR or FNR for 53 individual multicentre trials in critical care based on a pretest probability of 0.5 that the hypothesis was true. Results: For trials reporting statistical significance, the FPR varied between 0.1% and 57.6%. For trials reporting non-significance, the FNR varied between 1.7% and 36.9%. Twenty-six of 47 trials (55.3%) reporting non-significance provided strong or very strong evidence in favour of the null hypothesis; the remaining trials provided limited evidence. There was no obvious relationship between the P-value and the FNR. Conclusions: The FPR and FNR showed marked variability, indicating that the probability of a real or absent treatment effect differed substantially between trials. Only one trial reporting statistical significance provided convincing evidence of a real treatment effect, and nearly half of all trials reporting non-significance provided limited evidence for the absence of a treatment effect. Our findings suggest that the quality of evidence from multicentre trials in critical care is highly variable.

Language Can Obscure as Well as Facilitate Apparent-Theory of Mind Performance: Part 2-The Case of Dyslexia in Adulthood.

Many studies imply causal links between linguistic competencies and Theory of Mind (ToM). But despite Dyslexia being a prime example of linguistic deficits, studies on whether it is related to ToM have been relatively unforthcoming. In the first of 2 studies (N = 89), independently-diagnosed dyslexic adults and non-dyslexic adults were presented with false-belief vignettes via computer, answering 4 types of question (Factual, Inference, 1st-order ToM & 2nd-order ToM). Dyslexia related to lower false-belief scores. Study 2 (N = 93) replicated this result with a non-computer-based variant on the false-belief task. We considered the possibility that the apparent-issue with ToM is caused by processing demands more associated to domains of cognition such as language, than to ToM itself. Addressing this possibility, study 2 additionally utilised the ToM30Q questionnaire, designed largely to circumvent issues related to language and memory. Principal-Components analysis extracted 4 factors, 2 capturing perceptual/representational ToM, and the other 2 capturing affective components related to ToM. The ToM30Q was validated via its associations to a published measure of empathy, replication of the female gender advantage over males, and for one factor from the ToM30Q there was a correlation with an existing published index of ToM. However, when we considered the performance of dyslexic and non-dyslexic participants using the ToM30Q, we found absolutely no difference between them. The contrasting findings from our 2 studies here, arguably offer the first experimental evidence with adults, that there is in fact no ToM deficit in dyslexia. Additionally, this finding raises the possibility that some other groups considered in some sense atypical, failed ToM tasks, not because they actually have a ToM deficit at all, but rather because they are asked to reveal their ToM competence through cognitive domains, such as language and memory.

Towards a resolution of some outstanding issues in transitive research: An empirical test on middle childhood.

Transitive Inference (deduce B > D from B > C and C > D) can help us to understand other areas of sociocognitive development. Across three experiments, learning, memory, and the validity of two transitive paradigms were investigated. In Experiment 1 (N = 121), 7-year-olds completed a three-term nontraining task or a five-term task requiring extensive-training. Performance was superior on the three-term task. Experiment 2 presented 5-10-year-olds with a new five-term task, increasing learning opportunities without lengthening training (N = 71). Inferences improved, suggesting children can learn five-term series rapidly. Regarding memory, the minor (CD) premise was the best predictor of BD-inferential performance in both task-types. However, tasks exhibited different profiles according to associations between the major (BC) premise and BD inference, correlations between the premises, and the role of age. Experiment 3 (N = 227) helped rule out the possible objection that the above findings simply stemmed from three-term tasks with real objects being easier to solve than computer-tasks. It also confirmed that, unlike for five-term task (Experiments 1 & 2), inferences on three-term tasks improve with age, whether the age range is wide (Experiment 3) or narrow (Experiment 2). I conclude that the tasks indexed different routes within a dual-process conception of transitive reasoning: The five-term tasks indexes Type 1 (associative) processing, and the three-term task indexes Type 2 (analytic) processing. As well as demonstrating that both tasks are perfectly valid, these findings open up opportunities to use transitive tasks for educability, to investigate the role of transitivity in other domains of reasoning, and potentially to benefit the lived experiences of persons with developmental issues.

Clinical and manometric characteristics of patients with Parkinson's disease and esophageal symptoms.

Dysphagia is a common problem in patients with Parkinson's disease (PD); its etiology is multifactorial and its management is challenging. In this retrospective cohort analysis using prospectively collected data, we aimed to objectively characterize dysphagia and/or other esophageal symptoms in patients with PD, assess the prevalence of outflow obstruction as well as major or minor disorders of esophageal peristalsis leading to impaired esophageal clearance and highlight objective parameters that can help in the current management algorithm. Thirty-three consecutive patients with PD presenting with dysphagia, odynophagia, heartburn, regurgitation, chest pain, and weight loss underwent clinical and functional evaluation by high-resolution manometry (HRM). Esophagogastric junction (EGJ) outflow obstruction and major as well as minor disorders of peristalsis were then assessed using the Chicago classification (v3). Thirty-three PD patients with esophageal symptoms were enrolled in the study; 12 of them reported weight loss that was considered as potentially reflecting underlying esophageal dysfunction. The median age of the patients was 70 years (range: 53-89 years), 24 (75%) were men. The majority (62%) experienced dysphagia, likely contributing to weight loss in 41% of patients. Odynophagia was rare (6%) while GER symptoms, such as heartburn, regurgitation, and chest pain were noted in 37%, 31%, and 28% of patients, respectively. Using the hierarchy of the Chicago classification, 12 patients (39%) exhibited EGJ outflow obstruction, 16 (48%) diffuse esophageal spasm (DES), 18 (55%), ineffective esophageal peristalsis (IEM), 16 (48%) fragmented peristalsis, and only 2 patients (6%) had normal HRM tracings. There were no patients with HRM features of achalasia. Dysphagia is common in patients with PD and is associated with a high prevalence of underlying motility disturbances as identified by HRM. The exact impact of these motility abnormalities on symptom induction and their role in influencing clinical management are unclear and will require further study.

Effects of depression, metabolic syndrome, and cardiorespiratory fitness on mortality: results from the Cooper Center Longitudinal Study.

BACKGROUND: Depression and metabolic syndrome (MetS) are frequently comorbid disorders that are independently associated with premature mortality. Conversely, cardiorespiratory fitness (CRF) is associated with reduced mortality risk. These factors may interact to impact mortality; however, their effects have not been assessed concurrently. This analysis assessed the mortality risk of comorbid depression/MetS and the effect of CRF on mortality in those with depression/MetS. METHODS: Prospective study of 47 702 adults in the Cooper Center Longitudinal Study. Mortality status was attained from the National Death Index. History of depression was determined by patient response (yes or no) to a standardized medical history questionnaire. MetS was categorized using the American Heart Association/National Heart, Lung, and Blood Institute criteria. CRF was estimated from the final speed/grade of a treadmill graded exercise test. RESULTS: 13.9% reported a history of depression, 21.4% met criteria for MetS, and 3.0% met criteria for both MetS and history of depression. History of depression (HR = 1.24, p = 0.003) and MetS (HR = 1.28, p < 0.001) were independently associated with an increased mortality risk, with the greatest mortality risk among individuals with both a history of depression and MetS (HR = 1.59, p < 0.001). Higher CRF was associated with a significantly lower risk of mortality (p < 0.001) in all individuals, including those with MetS and/or a history of depression. CONCLUSIONS: Those with higher levels CRF had reduced mortality risk in the context of depression/MetS. Interventions that improve CRF could have substantial impact on the health of persons with depression/MetS.

What Stroop tasks can tell us about selective attention from childhood to adulthood.

A rich body of research concerns causes of Stroop effects plus applications of Stroop. However, several questions remain. We included assessment of errors with children and adults (N = 316), who sat either a task wherein each block employed only trials of one type (unmixed task) or where every block comprised of a mix of the congruent, neutral, and incongruent trials. Children responded slower than adults and made more errors on each task. Contrary to some previous studies, interference (the difference between neutral and incongruent condition) showed no reaction time (RT) differences by group or task, although there were differences in errors. By contrast, facilitation (the difference between neutral and congruent condition) was greater in children than adults, and greater on the unmixed task than the mixed task. After considering a number of theoretical accounts, we settle on the inadvertent word-reading hypothesis, whereby facilitation stems from children and the unmixed task promoting inadvertent reading particularly in the congruent condition. Stability of interference RT is explained by fixed semantic differences between neutral and incongruent conditions, for children versus adults and for unmixed versus mixed task. We conclude that utilizing two tasks together may reveal more about how attention is affected in other groups.

Measured removal rates of chrysotile asbestos fibers from air and comparison with theoretical estimates based on gravitational settling and dilution ventilation.

CONTEXT: Industrial hygiene assessments often focus on activity-based airborne asbestos concentration measurements, but few empirical data exist regarding the fiber removal rate from air after activities cease. OBJECTIVE: Grade 7T chrysotile indoor fiber settling (FS) rates were characterized using air sampling (NIOSH Method 7402). MATERIALS AND METHODS: Six replicate events were conducted in a 58 m(3) study chamber (ventilation 3.5 ACH), in which chrysotile-contaminated work clothing was manipulated for 15 min followed by 30 min of no activity. The fiber concentration decay constant and removal rate were characterized using an exponential decay model based on the measurements. RESULTS: Breathing zone airborne chrysotile concentrations decreased by 86% within 15-30 min after fiber disturbance, compared to concentrations during active disturbance (p < 0.05). Estimated mean time required for 99% of the phase contrast microscopy-equivalent (PCME) fibers to be removed from air was approximately 30 min (95% CI: 22-57 min). The observed effective FS velocity was 0.0034 m/s. This settling velocity was between 4.5-fold and 180-fold faster than predicted by two different particulate gravitational settling models. Additionally, PCME concentrations decreased approximately 2.5-fold faster than predicted due to air exchange alone (32 versus 79 min to 99% decrease in concentration). DISCUSSION: Other measurement studies have reported similar airborne fiber removal rates, supporting the finding that factors other than gravitational settling and dilution ventilation contribute measurably to PCM fiber removal from air (e.g. impaction, agglomeration). CONCLUSION: Overall, the scientific weight of evidence indicates that the time necessary for removal of 99% of fibers greater than 5 µm in length (with aspect ratios greater than 3:1) is approximately 20-80 min.

Factors and processes in children's transitive deductions.

Transitive tasks are important for understanding how children develop socio-cognitively. However, developmental research has been restricted largely to questions surrounding maturation. We asked 6-, 7- and 8-year-olds (N = 117) to solve a composite of five different transitive tasks. Tasks included conditions asking about item-C (associated with the marked relation) in addition to the usual case of asking only about item-A (associated with the unmarked relation). Here, children found resolving item-C much easier than resolving item-A, a finding running counter to long-standing assumptions about transitive reasoning. Considering gender perhaps for the first time, boys exhibited higher transitive scores than girls overall. Finally, analysing in the context of one recent and well-specified theory of spatial transitive reasoning, we generated the prediction that reporting the full series should be easier than deducing any one item from that series. This prediction was not upheld. We discuss amendments necessary to accommodate all our earlier findings.

Proneurogenic Group II mGluR antagonist improves learning and reduces anxiety in Alzheimer Aß oligomer mouse.

Proneurogenic compounds have recently shown promise in some mouse models of Alzheimer's pathology. Antagonists at Group II metabotropic glutamate receptors (Group II mGluR: mGlu2, mGlu3) are reported to stimulate neurogenesis. Agonists at those receptors trigger γ-secretase-inhibitor-sensitive biogenesis of Aß42 peptides from isolated synaptic terminals, which is selectively suppressed by antagonist pretreatment. We have assessed the therapeutic potential of chronic pharmacological inhibition of Group II mGluR in Dutch APP (Alzheimer's amyloid precursor protein E693Q) transgenic mice that accumulate Dutch amyloid-ß (Aß) oligomers but never develop Aß plaques. BCI-838 is a clinically well-tolerated, orally bioavailable, investigational prodrug that delivers to the brain BCI-632, the active Group II mGluR antagonist metabolite. Dutch Aß-oligomer-forming APP transgenic mice (APP E693Q) were dosed with BCI-838 for 3 months. Chronic treatment with BCI-838 was associated with reversal of transgene-related amnestic behavior, reduction in anxiety, reduction in levels of brain Aß monomers and oligomers, and stimulation of hippocampal neurogenesis. Group II mGluR inhibition may offer a unique package of relevant properties as an Alzheimer's disease therapeutic or prophylactic by providing both attenuation of neuropathology and stimulation of repair.

Demonstration of the need for cardiovascular and pulmonary normative data for cancer survivors.

Despite evidence that cancer and its treatments severely reduce cardiorespiratory fitness (CRF), normative data for cancer survivors do not exist. The present study identifies age and gender-specific CRF distributions in a cancer population. The use of cancer-specific normative CRF data may help stratify initial fitness status and assess improvements in response to exercise interventions in cancer survivors. Data from 703 cancer survivors were analyzed for this study. Quintiles were compiled for peak oxygen consumption (VO2peak), forced vital capacity (FVC), and forced expiratory volume (FEV1) for males and females in 5 age groups (19-39, 40-49, 50-59, 60-69, and ≥70 years of age). VO2peak values for the cancer population were significantly lower than the general US population. The cancer population average in each age group fell within the "very poor" classification of VO2peak values for the general population. FVC values in the cancer population were similar to the general population. Cancer survivors had very low age group-specific VO2peak values compared to the apparently healthy general US population. Previously, CRF values of cancer survivors were compared to normative values for the apparently healthy general population, which yielded imprecise classifications of initial fitness and changes in fitness, resulting in patient discouragement.

Evaluation of take-home exposure and risk associated with the handling of clothing contaminated with chrysotile asbestos.

The potential for para-occupational (or take-home) exposures from contaminated clothing has been recognized for the past 60 years. To better characterize the take-home asbestos exposure pathway, a study was performed to measure the relationship between airborne chrysotile concentrations in the workplace, the contamination of work clothing, and take-home exposures and risks. The study included air sampling during two activities: (1) contamination of work clothing by airborne chrysotile (i.e., loading the clothing), and (2) handling and shaking out of the clothes. The clothes were contaminated at three different target airborne chrysotile concentrations (0-0.1 fibers per cubic centimeter [f/cc], 1-2 f/cc, and 2-4 f/cc; two events each for 31-43 minutes; six events total). Arithmetic mean concentrations for the three target loading levels were 0.01 f/cc, 1.65 f/cc, and 2.84 f/cc (National Institute of Occupational Health and Safety [NIOSH] 7402). Following the loading events, six matched 30-minute clothes-handling and shake-out events were conducted, each including 15 minutes of active handling (15-minute means; 0.014-0.097 f/cc) and 15 additional minutes of no handling (30-minute means; 0.006-0.063 f/cc). Percentages of personal clothes-handling TWAs relative to clothes-loading TWAs were calculated for event pairs to characterize exposure potential during daily versus weekly clothes-handling activity. Airborne concentrations for the clothes handler were 0.2-1.4% (eight-hour TWA or daily ratio) and 0.03-0.27% (40-hour TWA or weekly ratio) of loading TWAs. Cumulative chrysotile doses for clothes handling at airborne concentrations tested were estimated to be consistent with lifetime cumulative chrysotile doses associated with ambient air exposure (range for take-home or ambient doses: 0.00044-0.105 f/cc year).

A teacher-centered exploration of the relevance of social factors to theory of mind development.

Many accounts of children's Theory of Mind (ToM) development favor a cognitive explanation, for example, in terms of mental representational improvements at or before 4 years. Here, we investigated whether social factors as rated by a child's teacher, are related to ToM development. We tested 82 children of 3-6 years on each of four ToM tasks, and their class teacher completed a social questionnaire about each child's playing behavior, sharing, talkativeness, confidence, aggressiveness and outgoingness. A measure of task memory and the child's gender were also recorded. Here, children generally passed ToM tasks after 5 years-old, but no one gender performed reliably better than the other. Teacher-rated confidence and playing behavior were correlated to ToM. But in a regression analysis, these were replaced by teacher-rated talkativeness; with age and memory given primacy in both sets of analyses. It is concluded that maturation and cognitive factors may well have primacy but social factors, facilitated during early primary education, must also be given a role in ToM development.

Management options for aorto-oesophageal fistula: case histories and review of the literature.

OBJECTIVE: An aorto-oesophageal fistula is a rare clinical entity, leading to life-threatening gastrointestinal bleeding. Thoracic aortic aneurysms are the most common cause of aorto-oesophageal fistulae; further causes involve foreign body ingestion, trauma (in most cases iatrogenic), carcinoma or, very rarely, aortitis tuberculotica. METHODS: Due to its rarity, there are no large multicentre studies present to evaluate the efficacy of different therapeutic management options. Since it is associated with significant morbidity and mortality, we give a short summary of various treatment approaches performed in our clinical practice in the past three years. The most straightforward therapeutic option may be an endovascular aortic repair and subtotal oesophageal resection followed by gastro-oesophageal reconstruction, but other alternative treatment possibilities are also present, although with probable higher morbidity. CONCLUSIONS: Eliminating the source of bleeding as an emergency, resecting the oesophagus urgently to prevent sepsis and reconstructing the gastrointestinal continuity as an elective case after having the inflammatory processes settled seems to justify the endovascular aortic repair and subtotal oesophageal resection, followed by a gastro-oesophageal reconstruction, as an effective surgical approach.

A child-centred exploration of the relevance of family and friends to theory of mind development.

Theory of Mind (ToM) is said to develop at around 4 years old. But some studies suggest it develops considerably earlier than this, with others suggesting it develops much later. Although several recent studies have found that social factors (like gender, family size, number of siblings, and number of friends) can impact on ToM, other studies contradict those findings. We wondered whether addressing several procedural issues and ensuring the task concerns real protagonists in real time, would bear on the above issues. Here, 114 children of 3-6 years completed four ToM tasks incorporating controls from experimental psychology, including randomly varying the order of ToM and non-ToM questions across participants. Now, children passed ToM tasks from around 5 years old, rather than 4 years or earlier. Girls did not develop ToM any earlier than boys. There was clear correlational evidence for the older-sibling effect and effects of friends but no reliable effects of nuclear or extended family. However, when these factors were set in the context of one another, the sibling effect was driven by a negative influence from younger siblings (as opposed to older siblings) and the friends effect was driven by friends at school (as opposed to friends at home). Finally, "friends" was a stronger predictor than siblings but memory (a cognitive factor) and age (a maturational factor) were the strongest predictors of all.

Retention rates and weight loss in a commercial weight loss program.

OBJECTIVE: To evaluate and describe retention rates and weight loss in clients participating in a commercial weight loss program. SUBJECTS: A total of 60 164 men and women ages 18-79 years who enrolled in the Jenny Craig Platinum program between May 2001 and May 2002. METHODS: Retention rates, mean weight loss and percent weight loss were calculated on a weekly basis for the 52-week period following initial enrollment in the weight loss program. Clients were categorized based on final week of participation in the program (weeks 1-4, weeks 5-13, weeks 14-26, weeks 27-39 and weeks 40-52) and weight loss was calculated at final week. A subgroup of clients was identified based on attendance through 13, 26 and 52 weeks. Mean and percent weight loss was calculated for these subgroups of clients. RESULTS: Of the 60 164 men and women who enrolled in the weight loss program, 73% were retained in the program after 4 weeks, 42% at 13 weeks, 22% at 26 weeks and 6.6% at 52 weeks. Clients who dropped out of the program during the first 4 weeks lost 1.1+/-1.6% (mean+/-s.d.) of their initial body weight, whereas clients who dropped out between 40 and 52 weeks lost 12.0+/-7.2%. Clients in the 13-week, 26-week and 52-week cohorts lost 8.3+/-3.3, 12.6+/-5.1 and 15.6+/-7.5% of their initial body weight, respectively. CONCLUSION: Weight loss was greater among clients who were retained in the program longer. The findings from this study suggest that a commercial weight loss program can be an effective weight loss tool for individuals who remain active in the program.

Emergency intensive care admission following elective thoracic surgery.

AIM: The aim of this study was to analyse the outcomes of patients admitted to the intensive care unit (ICU) following initial recovery after elective thoracic surgery. METHODS: The case notes of all patients who underwent elective thoracic surgery over a one-year period were reviewed. Patients who were admitted to ICU following an initial recovery on the ward were identified and their postoperative course analysed. The clinical and demographic characteristics of these patients were recorded and their outcomes analysed. RESULTS: A total of 20 patients were admitted to ICU of whom 13 (65%) were admitted for respiratory complication, 5 with sepsis and 2 with cardiovascular instability. Sixteen (80%) patients required CPAP or BIPAP, of whom only 7 (35%) required mechanical ventilation. Renal support was required in 7 patients, with 2 (10%) requiring haemofiltration. ICU survival was 15 patients (75%), whilst overall three-month survival post ICU admission was 65%. Requirement for renal support was the only predictor of mortality on univariate and multivariate analysis. CONCLUSIONS: Salvage ICU admission following elective thoracic surgery is associated with significant mortality, however the outcome is far from hopeless. The majority of patients can be managed without recourse to mechanical ventilation or haemofiltration. The need for renal support is, however, a significant adverse prognostic indicator.

Chlorambucil and lomustine (CL56) in absolute hormone refractory prostate cancer: re-induction of endocrine sensitivity an unexpected finding.

The management of androgen independent prostate cancer is increasingly disputed. Diethylstilbestrol and steroids have useful second-line activity in its management. The value of chemotherapy still remains contentious. This paper reports a phase 2 study of two orally active chemotherapy drugs in patients who are absolutely hormone refractory having failed primary androgen blockade and combined oestrogens and corticosteroids. In total, 37 patients who were biochemically castrate with absolute hormone refractory prostate cancer and performance status of 0-3 were enrolled. Therapy consisted of chlorambucil 1 mg kg(-1) given as 6 mg a day until the total dose was reached and lomustine 2 mg kg(-1) given every 56 days (CL56). During this time all hormone therapy was stopped. One patient normalised his PSA with a further two having a greater than 50% decline leading to an objective response rate of 10%. The median time to progression was 3.6 months with an overall survival of 7.1 months. The median survival of this group of patients from first becoming androgen independent was 23.5 months. Eight of 17 (47%) patients who were subsequently re-challenged with hormonal therapy following failure of chemotherapy had a further PSA reduction, three (17%) of which were >50%. The median progression-free interval for the eight patients was 4 months. In conclusion, CL56 has a low objective response rate in the management of absolute hormone refractory prostate cancer. Toxicity was mild. Re-induction of hormone sensitivity following failure of chemotherapy was an unexpected finding that requires further study.

Dose intense triplet chemotherapy with gemcitabine, carboplatin, paclitaxel with peripheral blood progenitor cell support for six cycles in advanced epithelial ovarian cancer.

The interval required for haematological reconstitution following myelosuppressive chemotherapy can be reduced by the infusion of autologous peripheral blood progenitor cells (PBPCs). When carboplatin (C) and paclitaxel (P) are followed by granulocyte colony-stimulating factor (GCSF), multiple courses can be given at 10-day intervals with the autologous PBPCs from a unit of whole blood with each cycle. We extended this approach and defined the dose-limiting toxicity and maximum-tolerated dose for the addition of gemcitabine (G) to CP for patients (pts) with EOC in a phase I-II study of increasing doses of G (0, 800, 1000 and 1250 mg x m(-2)) over four cohorts with C at area under curve (AUC) 6, plus P at 175 mg x m(-2) 3 h(-1) every 10 days for six cycles. Granulocyte colony-stimulating factor 5 microg x kg(-1) day(-1) was given s.c. days 1-10 and 450 ml whole blood was venesected before each treatment, stored untreated at 4 degrees C and reinfused 24 h later. In all, 17 patients with EOC either bulky stage IV or recurrent after treatment-free interval >12 months were treated over 30 months. Of the 17 patients, 13 completed six cycles (one patient stopped early with PD, three with toxicity), interdose interval 9-28 (median 10) days. Delays occurred in four patients due to infection or malaise, and there were no dose reductions. Haematological toxicity was not considered to be dose limiting. Febrile neutropenia was uncommon (2 patients), but grade III/IV thrombocytopenia was seen across all cohorts. Treatment was not delayed for thrombocytopenia and no bleeding complications occurred. Grade III transaminitis was seen in all patients in cohort 4 and grade IV toxicity, considered to be dose limiting, occurred in one. Responses were observed at all dose levels with six CR, seven PR, three SD and one PD. Dose intense GCP was deliverable over six cycles with manageable haematological toxicity, but with dose-limiting hepatic toxicity in cohort 4. The MTD was gemcitabine 1000 mg x m(-2), carboplatin AUC 6, paclitaxel 175 mg x m(-2) given every 10 days for six cycles.