Wrangling with p-values versus effect sizes to improve medical decision-making: A tutorial.

The most pervasive and damaging myth in clinical research is that the smaller the p-value, the stronger the hypothesis. In reality, the p-value primarily reflects the quality of research design decisions. The most common proposal to avoid misleading conclusions from clinical research requires the appropriate use of effect sizes, but which effect size, used when and how, is an open question. A solution is proposed for perhaps the most common problem in clinical research, the comparison between two populations, for example, comparison of two treatments in a randomized clinical trial or comparison of high risk versus low risk individuals in an epidemiological study: the success rate difference or equivalently the number needed to treat/take (NNT).

Depressive symptoms predict head and neck cancer survival: Examining plausible behavioral and biological pathways.

BACKGROUND: Head and neck cancers are associated with high rates of depression, which may increase the risk for poorer immediate and long-term outcomes. Here it was hypothesized that greater depressive symptoms would predict earlier mortality, and behavioral (treatment interruption) and biological (treatment response) mediators were examined. METHODS: Patients (n = 134) reported depressive symptomatology at treatment planning. Clinical data were reviewed at the 2-year follow-up. RESULTS: Greater depressive symptoms were associated with significantly shorter survival (hazard ratio, 0.868; 95% confidence interval [CI], 0.819-0.921; P < .001), higher rates of chemoradiation interruption (odds ratio, 0.865; 95% CI, 0.774-0.966; P = .010), and poorer treatment response (odds ratio, 0.879; 95% CI, 0.803-0.963; P = .005). The poorer treatment response partially explained the depression-survival relation. Other known prognostic indicators did not challenge these results. CONCLUSIONS: Depressive symptoms at the time of treatment planning predict overall 2-year mortality. Effects are partly influenced by the treatment response. Depression screening and intervention may be beneficial. Future studies should examine parallel biological pathways linking depression to cancer survival, including endocrine disruption and inflammation. Cancer 2018;124:1053-60. © 2018 American Cancer Society.

Young adult outcomes in the follow-up of the multimodal treatment study of attention-deficit/hyperactivity disorder: symptom persistence, source discrepancy, and height suppression.

BACKGROUND: The Multimodal Treatment Study (MTA) began as a 14-month randomized clinical trial of behavioral and pharmacological treatments of 579 children (7-10 years of age) diagnosed with attention-deficit/hyperactivity disorder (ADHD)-combined type. It transitioned into an observational long-term follow-up of 515 cases consented for continuation and 289 classmates (258 without ADHD) added as a local normative comparison group (LNCG), with assessments 2-16 years after baseline. METHODS: Primary (symptom severity) and secondary (adult height) outcomes in adulthood were specified. Treatment was monitored to age 18, and naturalistic subgroups were formed based on three patterns of long-term use of stimulant medication (Consistent, Inconsistent, and Negligible). For the follow-up, hypothesis-generating analyses were performed on outcomes in early adulthood (at 25 years of age). Planned comparisons were used to estimate ADHD-LNCG differences reflecting persistence of symptoms and naturalistic subgroup differences reflecting benefit (symptom reduction) and cost (height suppression) associated with extended use of medication. RESULTS: For ratings of symptom severity, the ADHD-LNCG comparison was statistically significant for the parent/self-report average (0.51 ± 0.04, p < .0001, d = 1.11), documenting symptom persistence, and for the parent/self-report difference (0.21 ± 0.04, p < .0001, d = .60), documenting source discrepancy, but the comparisons of naturalistic subgroups reflecting medication effects were not significant. For adult height, the ADHD group was 1.29 ± 0.55 cm shorter than the LNCG (p < .01, d = .21), and the comparisons of the naturalistic subgroups were significant: the treated group with the Consistent or Inconsistent pattern was 2.55 ± 0.73 cm shorter than the subgroup with the Negligible pattern (p < .0005, d = .42), and within the treated group, the subgroup with the Consistent pattern was 2.36 ± 1.13 cm shorter than the subgroup with the Inconsistent pattern (p < .04, d = .38). CONCLUSIONS: In the MTA follow-up into adulthood, the ADHD group showed symptom persistence compared to local norms from the LNCG. Within naturalistic subgroups of ADHD cases, extended use of medication was associated with suppression of adult height but not with reduction of symptom severity.

Higher vagal activity as related to survival in patients with advanced breast cancer: an analysis of autonomic dysregulation.

OBJECTIVE: High levels of high-frequency heart rate variability (HF-HRV), related to parasympathetic-nervous-system functioning, have been associated with longer survival in patients with myocardial infarction and acute trauma and in patients undergoing palliative care. From animal studies linking higher vagal activity with better immune system functioning and reduced metastases, we hypothesized that higher HF-HRV would predict longer survival in patients with metastatic or recurrent breast cancer (MRBC). METHODS: Eighty-seven patients with MRBC participated in a laboratory task including a 5-minute resting baseline electrocardiogram. HF-HRV was computed as the natural logarithm of the summed power spectral density of R-R intervals (0.15-0.50 Hz). In this secondary analysis of a study testing whether diurnal cortisol slope predicted survival, we tested the association between resting baseline HF-HRV on survival using Cox proportional hazards models. RESULTS: A total of 50 patients died during a median follow-up of 7.99 years. Higher baseline HF-HRV predicted significantly longer survival, with a hazard ratio of 0.75 (95% confidence interval = 0.60-0.92, p = .006). Visceral metastasis status and baseline heart rate were related to both HF-HRV and survival. However, a combination of HF-HRV and heart rate further improved survival prediction, with a hazard ratio of 0.64 (95% confidence interval = 0.48-0.85, p = .002). CONCLUSIONS: Vagal activity of patients with MRBC strongly predicted their survival, extending the known predictive window of HF-HRV in cancer beyond palliative care. Vagal activity can be altered by behavioral, pharmacological, and surgical interventions and may be a promising target for extending life expectancy in patients with metastasizing cancer.

Donepezil treatment in ethnically diverse patients with Alzheimer disease.

OBJECTIVE: To compare the outcome of donepezil treatment in ethnically diverse Alzheimer disease (AD) patients with ethnically diverse AD patients who did not receive donepezil. METHODS: Patients meeting NINCDS-ADRA criteria for probable or possible AD from a consortium of California sites were systematically followed for at least 1 year in this prospective, observational study. Their treatment regimens, including prescription of donepezil, were determined by their individual physician according to his or her usual criteria. Patients self-identified their ethnicity. RESULTS: The 64 ethnically diverse AD patients who completed the study and received donepezil treatment had an average 1-year decline of 2.30 points (standard deviation: 3.9) on the 30-point Mini-Mental State Exam compared with a 1.70-point (standard deviation: 4.2) decline in the 74 ethnically diverse completers who received no donepezil or other anti-AD drugs during the study period. This difference was not statistically significant. The overall Cohen effect size of this treatment-associated difference was estimated at -0.15. After using propensity analyses and other techniques to assess factors that could bias prescribing decisions, the lack of benefits associated with donepezil treatment remained. The lack of donepezil benefits also remained when more traditional analyses were applied to these data. CONCLUSION: Ethnically diverse AD patients in this study apparently did not benefit from 1 year of donepezil treatment. These unpromising results are in contrast to modest benefits of donepezil treatment measured in a directly comparable California study involving white non-Latino AD patients.

Cognitive markers for efficacy of neurofeedback for attention-deficit hyperactivity disorder - personalized medicine using computational psychiatry in a randomized clinical trial.

BACKGROUND: Exploring whether cognitive components (identified by baseline cognitive testing and computational modeling) moderate clinical outcome of neurofeedback (NF) for attention-deficit hyperactivity disorder (ADHD). METHOD: 142 children (aged 7-10) with ADHD were randomly assigned to either NF (n = 84) or control treatment (n = 58) in a double-blind clinical trial (NCT02251743). The NF group received live, self-controlled downtraining of electroencephalographic theta/beta ratio power. The control group received identical-appearing reinforcement from prerecorded electroencephalograms from other children. 133 (78 NF, 55 control) children had cognitive processing measured at baseline with the Integrated Visual and Auditory Continuous Performance Test (IVA2-CPT) and were included in this analysis. A diffusion decision model applied to the IVA2-CPT data quantified two latent cognitive components deficient in ADHD: drift rate and drift bias, indexing efficiency and context sensitivity of cognitive processes involving information integration. We explored whether these cognitive components moderated the improvement in parent- and teacher-rated inattention symptoms from baseline to treatment end (primary clinical outcome). RESULTS: Baseline cognitive components reflecting information integration (drift rate, drift bias) moderated the improvement in inattention due to NF vs. control treatment (p = 0.006). Specifically, those with either the most or least severe deficits in these components showed more improvement in parent- and teacher-rated inattention when assigned to NF (Cohen's d = 0.59) than when assigned to control (Cohen's d = -0.21). CONCLUSIONS: Pre-treatment cognitive testing with computational modeling identified children who benefitted more from neurofeedback than control treatment for ADHD.

The future of psychopharmacology: a critical appraisal of ongoing phase 2/3 trials, and of some current trends aiming to de-risk trial programmes of novel agents.

Despite considerable progress in pharmacotherapy over the past seven decades, many mental disorders remain insufficiently treated. This situation is in part due to the limited knowledge of the pathophysiology of these disorders and the lack of biological markers to stratify and individualize patient selection, but also to a still restricted number of mechanisms of action being targeted in monotherapy or combination/augmentation treatment, as well as to a variety of challenges threatening the successful development and testing of new drugs. In this paper, we first provide an overview of the most promising drugs with innovative mechanisms of action that are undergoing phase 2 or 3 testing for schizophrenia, bipolar disorder, major depressive disorder, anxiety and trauma-related disorders, substance use disorders, and dementia. Promising repurposing of established medications for new psychiatric indications, as well as variations in the modulation of dopamine, noradrenaline and serotonin receptor functioning, are also considered. We then critically discuss the clinical trial parameters that need to be considered in depth when developing and testing new pharmacological agents for the treatment of mental disorders. Hurdles and perils threatening success of new drug development and testing include inadequacy and imprecision of inclusion/exclusion criteria and ratings, sub-optimally suited clinical trial participants, multiple factors contributing to a large/increasing placebo effect, and problems with statistical analyses. This information should be considered in order to de-risk trial programmes of novel agents or known agents for novel psychiatric indications, increasing their chances of success.

Health-service use in women with binge eating disorders.

OBJECTIVE: To compare health-care utilization between participants who met DSM-IV criteria for binge eating disorder (BED) and those engaged in recurrent binge eating (RBE) and to evaluate whether objective binge eating (OBE) days, a key measurement for diagnosing BED, predicted health-care costs. METHOD: We obtained utilization and cost data from electronic medical records to augment patient reported data for 100 adult female members of a large health maintenance organization who were enrolled in a randomized clinical trial to treat binge eating. RESULTS: Total costs did not differ between the BED and RBE groups (ß = -0.117, z = -0.48, p = .629), nor did the number of OBE days predict total costs (ß = -0.017, z = -1.01, p = .313). DISCUSSION: Findings suggest that the medical impairment, as assessed through health care costs, caused by BED may not be greater than impairment caused by RBE. The current threshold number of two OBE days/week as a criterion for BED may need to be reconsidered.

Screening for binge eating disorders using the Patient Health Questionnaire in a community sample.

OBJECTIVE: To examine the operating characteristics of the Patient Health Questionnaire eating disorder module (PHQ-ED) for identifying bulimia nervosa/binge eating disorder (BN/BED) or recurrent binge eating (RBE) in a community sample and to compare true positive (TP) versus false positive (FP) cases on clinical validators. METHOD: Two hundred and fifty-nine screen-positive individuals and a random sample of 89 screen negative cases completed a diagnostic interview. Sensitivity, specificity, and positive predictive value (PPV) were calculated. TP and FP cases were compared using t-tests and Chi-square tests. RESULTS: The PHQ-ED had high sensitivity (100%) and specificity (92%) for detecting BN/BED or RBE, but PPV was low (10 or 19%). TP and FP cases did not differ significantly on frequency of subjective bulimic episodes, objective overeating, restraint, on BMI, and on self-rated health. DISCUSSION: The PHQ-ED is recommended for use in large populations only in conjunction with follow-up questions to rule out cases without objective bulimic episodes.

Common Types of Gender-Based Microaggressions in Medicine.

PURPOSE: Microaggressions are subtle verbal or nonverbal everyday behaviors that arise from unconscious bias, covert prejudice, or hostility. They may contribute to the persistent disparities faced by women in medicine. In this study, the authors sought to identify common microaggressions experienced by women faculty in medicine and to determine if specific demographic characteristics affect the reported frequencies of these microaggressions. METHOD: The authors used chain referral sampling to collect real-life anecdotes about microaggressions from women faculty across the nation. Thirty-four unique experiences from those reported were identified and scripted then reenacted using professional actors to create 34 videos of the real-life microaggressions and 34 corresponding fictional "control" versions of the same situations. The videos, presented in a random order, were evaluated by faculty from 4 academic medical centers from 2016 to 2018. RESULTS: A total of 124 faculty (79 women, 45 men) participated. Women reported higher frequencies of microaggressions than men in 33 of the 34 videos depicting microaggressions (P value range: < .001 to .042, area under the curve range: 0.60-0.69). No such differences were seen with the control videos. Women identified 21 microaggressions as occurring frequently. No significant differences were found with respect to participants' age, race/ethnicity, academic rank, or years in medicine. Post hoc analyses showed that the microaggressions fell into 6 themes: encountering sexism, encountering pregnancy- and child care-related bias, having abilities underestimated, encountering sexually inappropriate comments, being relegated to mundane tasks, and feeling excluded/marginalized. CONCLUSIONS: Privilege is often invisible to those who have it, whereas bias and discrimination are readily apparent to those who experience it. Knowledge of common microaggressions will allow for targeted individual, interpersonal, and institutional solutions to mitigate disparities in medicine.

Use and misuse of the consolidated standards of reporting trials (CONSORT) guidelines to assess research findings: comment on Coyne, Stefanek, and Palmer (2007).

The basic principles underlying randomized clinical trials have been known for more than 50 years. The Consolidated Standards of Reporting Trials (CONSORT) guidelines, published in 1996 and based on those principles, are a valuable guide to what needs to be reported from any trial within word-limit constraints, but they do not provide guidelines to the decisions that have to be made to generate a trial with credible results. Using these guidelines as do J. C. Coyne, M. Stefanek, and S. C. Palmer (2007) is a misinterpretation of their purpose. Furthermore, Coyne et al. misrepresented the methods and findings of studies of the effects of psychosocial intervention on cancer survival. These errors are systematically reviewed. The results of these questionable analyses led Coyne et al. to recommend stopping research in an area that may be highly productive. Recent developments in the field are summarized. It is a mistaken and dangerous conclusion to declare this or other areas of scientific research off limits.

Neuroanatomy of fragile X syndrome is associated with aberrant behavior and the fragile X mental retardation protein (FMRP).

OBJECTIVE: To determine how neuroanatomic variation in children and adolescents with fragile X syndrome is linked to reduced levels of the fragile X mental retardation-1 protein and to aberrant cognition and behavior. METHODS: This study included 84 children and adolescents with the fragile X full mutation and 72 typically developing control subjects matched for age and sex. Brain morphology was assessed with volumetric, voxel-based, and surface-based modeling approaches. Intelligence quotient was evaluated with standard cognitive testing, whereas abnormal behaviors were measured with the Autism Behavior Checklist and the Aberrant Behavior Checklist. RESULTS: Significantly increased size of the caudate nucleus and decreased size of the posterior cerebellar vermis, amygdala, and superior temporal gyrus were present in the fragile X group. Subjects with fragile X also demonstrated an abnormal profile of cortical lobe volumes. A receiver operating characteristic analysis identified the combination of a large caudate with small posterior cerebellar vermis, amygdala, and superior temporal gyrus as distinguishing children with fragile X from control subjects with a high level of sensitivity and specificity. Large caudate and small posterior cerebellar vermis were associated with lower fragile X mental retardation protein levels and more pronounced cognitive deficits and aberrant behaviors. INTERPRETATION: Abnormal development of specific brain regions characterizes a neuroanatomic phenotype associated with fragile X syndrome and may mediate the effects of FMR1 gene mutations on the cognitive and behavioral features of the disorder. Fragile X syndrome provides a model for elucidating critical linkages among gene, brain, and cognition in children with serious neurodevelopmental disorders.

A dimensional versus a categorical approach to diagnosis: anxiety and depression in the HUNT 2 study.

The aim of this study was to compare a dimensional and a categorical approach to diagnosis, using as an illustration co-occurring symptoms of anxiety and depression concerning description, associations and predictive power. We analysed data from 60 869 individuals with valid ratings on the Hospital Anxiety and Depression Scale (HADS) and on mental impairment in the age range of 20 to 89 years of the cross-sectional Nord-Trøndelag Health Study 1995-1997. There was a wide variation of the dimensional symptom level (subscale scores) within both diagnostic categories (cut-offs > > or = 8 on both subscales), as is usually true with categorical and dimensional diagnosis. The dimensional (Spearman) correlation coefficients between anxiety and depression was 0.51 compared to 0.38 for the categorical. The power to predict impairment was weaker with the categorical than with the dimensional approach of the HADS, showing fewer statistically significant coefficients in the logistic regression models and lower area under curve (0.82 versus 0.87). This is an example illustrating the impact use of dimensional diagnoses would have on research and clinical practice.

Physical and mental health-related quality of life among older people with schizophrenia.

OBJECTIVE: Since the time of Kraeplin, schizophrenia has been thought of as a disorder with progressive deterioration in functioning. An important aspect of functioning is both physical and mental health-related quality of life (HRQoL). The objective of this study was to examine the relationship of age to both mental and physical aspects of HRQoL in individuals with schizophrenia as compared to normal comparison subjects (NCs). METHODS: Middle-aged and older community-dwelling patients with schizophrenia (N=486) were compared to NCs (N=101). Health related quality of life was measured using the SF-36 Physical Health and Mental Health Component scores. The relationship between age and HRQoL was examined using linear regressions. In addition, we performed exploratory analyses to examine the effects of confounding variables on this relationship, and to examine the effects of age on SF-36 subscales. RESULTS: Patients with schizophrenia had lower SF-36 Physical and Mental Health Component scores than NCs, and these differences persisted after adjusting for the age difference between the two groups. The relationship between age and mental, but not physical, HRQoL was significantly different between the patients with schizophrenia and the NCs. Specifically, older age was associated with higher mental HRQoL among patients with schizophrenia, but not among the NCs. This difference remained significant after examining multiple potential confounding demographic and clinical variables. CONCLUSIONS: This study found that older age was associated with greater mental health quality of life. Longitudinal studies are warranted to confirm our finding, and to examine potential mechanisms responsible for possible improvement in mental HRQoL with age.

Screening for childhood mental health problems: outcomes and early identification.

BACKGROUND: Many childhood psychiatric problems are transient. Consequently, screening procedures to accurately identify children with problems unlikely to remit and thus, in need of intervention, are of major public health concern. This study aimed to develop a universal school-based screening procedure based on the answers to three questions: (1) What are the broad patterns of mental health problems from kindergarten to grade 5? (2) What are the grade 5 outcomes of these patterns? (3) How early in school can children likely to develop the most impairing patterns be identified accurately? METHODS: Mothers and teachers reported on a community sample (N = 328) of children's internalizing and externalizing symptoms in kindergarten and grades 1, 3, and 5. In grade 5, teachers reported on children's school-based functional impairments, physical health problems, and service use; mothers reported on children's specialty mental health care. RESULTS: Four patterns distinguished children who (1) never evidenced symptoms; (2) evidenced only isolated symptoms; or evidenced recurrent symptoms, either (3) without or (4) with comorbid internalizing and externalizing. By grade 5, children with recurrent comorbid symptoms had the greatest impairments, physical health problems, and service use. These children can be identified quite accurately by grade 1. CONCLUSIONS: Universal screening at school entry can effectively identify children likely to develop recurrent comorbid symptoms, and would provide a basis for developing optimal targeted intervention programs.

Gender difference in the prevalence of eating disorder symptoms.

OBJECTIVE: This study examined gender differences in prevalence of eating disorder symptoms including body image concerns (body checking or avoidance), binge eating, and inappropriate compensatory behaviors. METHOD: A random sample of members (ages 18-35 years) of a health maintenance organization was recruited to complete a survey by mail or on-line. Items were drawn from the Patient Health Questionnaire and the Body Shape Questionnaire. RESULTS: Among the 3,714 women and 1,808 men who responded, men were more likely to report overeating, whereas women were more likely to endorse loss of control while eating. Although statistically significant gender differences were observed, with women significantly more likely than men to report body checking and avoidance, binge eating, fasting, and vomiting, effect sizes ("Number Needed to Treat") were small to moderate. DISCUSSION: Few studies of eating disorders include men, yet our findings suggest that a substantial minority of men also report eating disorder symptoms.

Sex-specific predictors of criminal recidivism in a representative sample of incarcerated youth.

OBJECTIVE: The objective of the study was to identify sex-specific psychopathologic predictors of criminal recidivism among a representative sample of incarcerated youths. METHOD: In this prospective longitudinal study, the Mini-International Psychiatric Interview for children and adolescents was used to assess psychopathology in juveniles entering an Austrian pretrial detention facility between March 2003 and January 2005. From the beginning of the study until January 2006, data on criminal history were obtained from the Integrierte Vollzugsverwaltung, a database containing criminal information of every individual incarcerated in Austria. Of the 370 eligible participants, the final study sample comprised 328 juveniles (56 girls and 272 boys, age range = 14-21 years, mean = 16.7). RESULTS: Reincarceration rates within the specified follow-up period were 52.6% for the boys and 37.5% for the girls. Using Cox forward stepwise regression and Kaplan-Meier analyses, age at first incarceration (B = -.296, Wald statistic = 17.11, P < .001) and oppositional defiant disorder (B = .751, Wald statistic = 19.25, P < .001) were identified as significant predictors for reoffending in boys. In girls, generalized anxiety disorder (B = 1.97, Wald statistic = 13.71, P < .001) was found to be a predictor for reoffending, whereas dysthymia (B = -1.44, Wald statistic = 4.02, P = .045) was found to serve as protective factor. CONCLUSION: Our study confirms high rates of reoffending after release from correctional facilities in both sexes. It further defines sex-specific psychopathologic risk factors for relapse in incarcerated juveniles. According to our results, in boys, oppositional defiant disorder and early age at first incarceration are predictive of reincarceration. In girls, anxiety disorder was found to be a risk factor for future offending, whereas dysthymia was found to have a protective influence. Consequently, rehabilitation programs should be sex specific.

Psychosocial predictors of resilience after the September 11, 2001 terrorist attacks.

The terrorist attacks of September 11, 2001 inflicted distress beyond those directly exposed, thereby providing an opportunity to examine the contributions of a range of factors (cognitive, emotional, social support, coping) to psychological resilience for those indirectly exposed. In an Internet convenience sample of 1281, indices of resilience (higher well-being, lower distress) at baseline (2.5-12 weeks post-attack) were each associated with less emotional suppression, denial and self-blame, and fewer negative worldview changes. After controlling for initial outcomes, baseline negative worldview changes and aspects of social support and coping all remained significant predictors of 6-month outcomes, with worldview changes bearing the strongest relationship to each. These findings highlight the role of emotional, coping, social support, and particularly, cognitive variables in adjustment after terrorism.

Sexual Adjustment and Body Image Scale (SABIS): a new measure for breast cancer patients.

The purpose of this study was to develop and validate a self-report measure of body image and sexual adjustment in breast cancer patients: the Sexual Adjustment and Body Image Scale (SABIS). Three hundred and fifty three women diagnosed with primary breast cancer that had completed initial surgical treatment completed the SABIS and five measures of psychological, psychosocial, and sexual functioning. Psychometric properties of the SABIS were examined and it was found to be a reliable and valid means of assessing body image and sexuality in breast cancer patients following surgery.

"Not just another meta-analysis": Sources of heterogeneity in psychosocial treatment effect on cancer survival.

BACKGROUND: Currently, there are eight meta-analyses that address the question whether psychosocial intervention can prolong survival with widely disparate conclusions. One reason for inconsistent findings may be the methods by which previous meta-analyses were conducted. METHODS: Databases were searched to identify valid randomized controlled trials that compared psychosocial intervention with usual care. Hazard ratios (HRs) and their confidence intervals were pooled to estimate the strength of the treatment effect on survival time, and z-tests were performed to assess possible heterogeneity of effect sizes associated with different patient and treatment characteristics. RESULTS: Twelve trials involving 2439 cancer patients that met screening criteria were included. The overall effect favored the treatment group with a HR of 0.71 (95% Cl 0.58-0.88; P = 0.002). An effect size favoring treatment group was observed in studies sampling lower vs higher percentage of married patients' (NNT = 4.3 vs NNT = 15.4), when Cognitive-Behavioral Therapy was applied at early vs late cancer stage (NNT = 2.3 vs NNT = -28.6), and among patients' older vs younger than 50 (NNT = 4.2 vs NNT = -20.5). CONCLUSIONS: Psychosocial interventions may have an important effect on survival. Reviewed interventions appear to be more effective in unmarried patients, patients who are older, and those with an early cancer stage who attend CBT. Limitations of previous meta-analysis are discussed.