Journal of Clinical Epidemiology RSS feed: Current Issue. We aim at promoting the quality of clinical and patient-oriented health services research through the advancement and application of innovative methods of: conducting and presenting primary research; synthesizing research results; disseminating results; and translating results into optimal clinical practice; with special attention to the training of new generations of scientists and clinical practice leaders Subscribers can now access the full-text online version of the Journal of Clinical Epidemiology . Journal of Clinical Epidemiology is ranked 21st of 105 journals in the Public, Environmental & Occupational Health category on the 2009 Journal Citation Reports®, published by Thomson Reuters, and has an Impact Factor of 2.896.http://www.jclinepi.com/?rss=yesElsevier Inc.en © 2010 Published by Elsevier Inc. All rights reserved. Journal of Clinical Epidemiology0895-4356633March 2010 © 2010 Published by Elsevier Inc. All rights reserved. healthpermissions@elsevier.comhttp://www.jclinepi.com/article/PIIS089543561000003X/abstract?rss=yesEditorial Board10.1016/S0895-4356(10)00003-XJournal of Clinical Epidemiology 63, 3 (2010)2010-03-01Journal of Clinical Epidemiology2010-03-01633S0895-4356(10)X0002-6IFCIFChttp://www.jclinepi.com/article/PIIS0895435610000065/abstract?rss=yesTable of Contents10.1016/S0895-4356(10)00006-5Journal of Clinical Epidemiology 63, 3 (2010)2010-03-01Journal of Clinical Epidemiology2010-03-01633S0895-4356(10)X0002-6iiiivhttp://www.jclinepi.com/article/PIIS0895435609003904/abstract?rss=yesIn the Meta-analyses and Systematic Review section of this issue, the question addressed is, “Why is individual patient data so rarely used in systematic reviews?” Individual patient data analysis is only used in 2% of published systematic reviews. As van Walraven points out in his commentary, individual patient data have many substantive and methodological advantages including outcome harmonization, analytic harmonization, and exploration of effectiveness variability. The Equity Group of the Campbell and Cochrane Collaborations (www.equity.cochrane.org), for which one of the editors is co-convener, is now encouraging systematic reviewers to analyze the effectiveness differences within different types of disadvantaged groups (e.g., poor, uneducated, out of work, place of residence, social isolation, and gender); this really needs individual patient data to classify each person and their outcome by their disadvantaged status. One of the reasons for rarely using individual patient data is probably that the clinical epidemiology community has neither aggressively showcased good examples when teaching evidence-based medicine and critical appraisal to the clinician readers nor have we emphasized this when teaching in graduate programs and in theses; we should review this approach.Advantages of individual patient data analysis in systematic reviewsPeter Tugwell, J. Andre Knottnerus10.1016/j.jclinepi.2009.12.005Journal of Clinical Epidemiology 63, 3 (2010)2010-03-01Journal of Clinical Epidemiology2010-03-01633S0895-4356(10)X0002-6Editorial233234http://www.jclinepi.com/article/PIIS0895435609001103/abstract?rss=yesMeta-analysis is an essential tool for summarizing medical research and determining the efficacy of therapies and procedures. Meta-analysis has been used in all areas of medical practice and aims to produce an overall estimate of the average treatment effect. Traditionally, this has been accomplished using information and results from published studies. Although such aggregate data meta-analyses (ADMA) are occasionally supplemented with additional data from study investigators, patient information in ADMA is always aggregated to the study level.Individual patient meta-analysis—rewards and challengesCarl van Walraven10.1016/j.jclinepi.2009.04.001Journal of Clinical Epidemiology 63, 3 (2010)2009-07-13Journal of Clinical Epidemiology2009-07-13633S0895-4356(10)X0002-6Commentary235237http://www.jclinepi.com/article/PIIS0895435609001279/abstract?rss=yesAbstract: Objective: To compare the results of meta-analysis of nonrandomized comparative studies (NRCSs) of a surgical procedure with that of randomized controlled trials (RCTs), and to assess the effect of design and conduct issues in NRCSs on measured outcomes.Study Design and Setting: Two meta-analyses of RCTs and NRCSs (2,512 and 6,438 procedures, respectively) of laparoscopic resection for colorectal cancer were performed according to accepted protocols, and 13 outcomes common between them were compared. Odds ratios (ORs) and 95% confidence intervals (CI) for dichotomous outcomes were assessed for the degree of overlap. Continuous outcomes were compared using cumulative weighted ratios (CWRs) and percentages for which a mean and standard deviation (SD) were calculated. The effects of design and conduct issues in the meta-analysis of NRCSs on measured morbidity rates were assessed using subgroup analysis.Results: The ORs of the three dichotomous outcomes overlapped widely. For the 10 continuous variables, the mean difference (SD) in the results of the two meta-analyses was only 5.6% (4.9%). Fulfillment of certain quality and conduct issues in the NRCSs determined the statistical homogeneity of the results of meta-analysis and their comparability with the “gold standard.”Conclusion: Meta-analysis of well-designed NRCSs of surgical procedures is probably as accurate as that of RCTs.Meta-analysis of well-designed nonrandomized comparative studies of surgical procedures is as good as randomized controlled trialsNed S. Abraham, Christopher J. Byrne, Jane M. Young, Michael J. Solomon10.1016/j.jclinepi.2009.04.005Journal of Clinical Epidemiology 63, 3 (2010)2009-08-28Journal of Clinical Epidemiology2009-08-28633S0895-4356(10)X0002-6Systematic Reviews and Meta Analysis238245http://www.jclinepi.com/article/PIIS0895435609002789/abstract?rss=yesAbstract: Objective: Conclusions based on meta-analyses of randomized trials carry a status of “truth.” Methodological components may identify trials with systematic errors (“bias”). Trial sequential analysis (TSA) evaluates random errors in meta-analysis. We analyzed meta-analyses on laparoscopic vs. small-incision cholecystectomy regarding different outcome measures for the occurrence of type I errors.Study Design and Setting: Using TSA, we calculated the required information size (IS) and the trial sequential monitoring boundaries regarding complications in our Cochrane review with meta-analyses of cholecystectomy. For each outcome, we calculated a low risk of bias heterogeneity-adjusted IS. As a sensitivity analysis, we calculated an a priori heterogeneity-adjusted IS.Results: According to the trial sequential analyses based on a low risk of bias heterogeneity-adjusted IS definitive evidence may be reached by conducting one more randomized trial. Information may be required on 582 and 119 additional randomized patients to evaluate the effect on severe complications and serious adverse events (SAEs), respectively.Conclusion: Our results provide incentives to conduct a new trial with a low risk of bias focusing on a new composite outcome measure of SAEs to obtain conclusive evidence on which operative method to recommend.Trial sequential analyses of meta-analyses of complications in laparoscopic vs. small-incision cholecystectomy: more randomized patients are neededFrederik Keus, Jørn Wetterslev, Christian Gluud, Hein G. Gooszen, Cornelis J.H.M. van Laarhoven10.1016/j.jclinepi.2009.08.023Journal of Clinical Epidemiology 63, 3 (2010)2009-12-11Journal of Clinical Epidemiology2009-12-11633S0895-4356(10)X0002-6Systematic Reviews and Meta Analysis246256http://www.jclinepi.com/article/PIIS0895435609001292/abstract?rss=yesAbstract: Objective: Because of the lack of a gold standard, the diagnostic performance of tests for the detection of latent tuberculosis infection (LTBI) is not known. However, statistical methods can be used to estimate the accuracy from the studies reporting the concordance among the tests.Study Design and Setting: We developed a random-effect latent-class model to estimate performance characteristics of three LTBI diagnostic tests: tuberculin skin test (TST, at 10-mm cutoff), QuantiFERON-TB gold (QFG), and TSPOT-TB from the studies evaluating agreement among the tests.Results: Nineteen studies were included. QFG had a sensitivity of 0.642 (95% confidence interval [CI]: 0.593–0.691) and specificity of 0.996 (95% CI: 0.989–1.000), TSPOT-TB had a sensitivity of 0.500 (95% CI: 0.334–0.666) and specificity of 0.906 (95% CI: 0.882–0.929), and TST had a sensitivity of 0.709 (95% CI: 0.658–0.761) and specificity of 0.683 (95% CI: 0.522–0.844). Results were not sensitive to the inclusion of any single study. When only the three studies that reported on TSPOT were removed, estimates for the other two tests varied minimally.Conclusions: Statistical methods can help estimate the accuracy of LTBI tests. Although the specificities were close to their reported values in the literature, the estimates for sensitivities were low; a finding that should be carefully evaluated.A statistical method was used for the meta-analysis of tests for latent TB in the absence of a gold standard, combining random-effect and latent-class methods to estimate test accuracyMohsen Sadatsafavi, Neal Shahidi, Fawziah Marra, Mark J. FitzGerald, Kevin R. Elwood, Na Guo, Carlo A. Marra10.1016/j.jclinepi.2009.04.008Journal of Clinical Epidemiology 63, 3 (2010)2009-08-19Journal of Clinical Epidemiology2009-08-19633S0895-4356(10)X0002-6Systematic Reviews and Meta Analysis257269http://www.jclinepi.com/article/PIIS0895435609001498/abstract?rss=yesAbstract: Objective: To compare Cohen's guidelines for small (0.2), medium (0.5), and large (0.8) effect sizes with empirical estimates for a cancer-specific health-related quality-of-life questionnaire (HRQOL), the Functional Assessment of Cancer Therapy - General (FACT-G).Methods: Seventy-one papers satisfied inclusion criteria for meta-analysis. Blinded to the HRQOL results, three “experts” (with expertise in interpreting the FACT-G questionnaire and managing cancer patients), predicted the relative magnitude of HRQOL mean differences. Size classes (small, medium, large) were defined in terms of relevance to clinical decision making. The experts worked independently and based their predictions on patient characteristics and clinical circumstances. Their judgments were linked with FACT-G results and inverse-variance–weighted mean effect sizes calculated for each size class.Results: At least two experts were perfectly concordant and up to one was discordant by at most one size category for 833 of the mean differences; for these, weighted kappas were generally in the “substantial” range (0.60–0.79). Of these mean differences, 617 were cross-sectional; small, medium, and large mean effect sizes were physical well-being 0.42, 0.87, 1.6; functional well-being 0.37, 0.71, 1.6; emotional well-being 0.32, 0.40, no large differences; and social well-being 0.14, 0.23, no large differences. Two hundred and sixteen longitudinal mean differences yielded small and medium effect sizes: physical well-being 0.26, 0.34; functional well-being 0.14, 0.28; emotional well-being 0.27, 0.23; and social well-being 0.08, 0.01. There was virtually no evidence for large longitudinal effects.Conclusion: These results provide specific, evidence-based alternatives to Cohen's generic guidelines, for use in sample-size calculations for the FACT-G and interpretation of the clinical significance of effects measured with FACT-G.Meta-analysis provides evidence-based effect sizes for a cancer-specific quality-of-life questionnaire, the FACT-GMadeleine T. King, Martin R. Stockler, David F. Cella, David Osoba, David T. Eton, Joanna Thompson, Amy R. Eisenstein10.1016/j.jclinepi.2009.05.001Journal of Clinical Epidemiology 63, 3 (2010)2009-08-28Journal of Clinical Epidemiology2009-08-28633S0895-4356(10)X0002-6Systematic Reviews and Meta Analysis270281http://www.jclinepi.com/article/PIIS0895435609002236/abstract?rss=yesAbstract: Objective: Meta-analysis yields a biased result if published studies represent a biased selection of the evidence. Copas proposed a selection model to assess the sensitivity of meta-analysis conclusions to possible selection bias. An alternative proposal is the trim-and-fill method. This article reports an empirical comparison of the two methods.Study Design and Setting: We took 157 meta-analyses with binary outcomes, analyzed each one using both methods, then performed an automated comparison of the results. We compared the treatment estimates, standard errors, associated P-values, and number of missing studies estimated by both methods.Results: Both methods give similar point estimates, but standard errors and P-values are systematically larger for the trim-and-fill method. Furthermore, P-values from the trim-and-fill method are typically larger than those from the usual random effects model when no selection bias is detected. By contrast, P-values from the Copas selection model and the usual random effects model are similar in this setting. The trim-and-fill method reports more missing studies than the Copas selection model, unless selection bias is detected when the position is reversed.Conclusions: The assumption that the most extreme studies are missing leads to excessively conservative inference in practice for the trim-and-fill method. The Copas selection model appears to be the preferable approach.Empirical evaluation suggests Copas selection model preferable to trim-and-fill method for selection bias in meta-analysisGuido Schwarzer, James Carpenter, Gerta Rücker10.1016/j.jclinepi.2009.05.008Journal of Clinical Epidemiology 63, 3 (2010)2009-10-19Journal of Clinical Epidemiology2009-10-19633S0895-4356(10)X0002-6Systematic Reviews and Meta Analysis282288http://www.jclinepi.com/article/PIIS0895435609001280/abstract?rss=yesAbstract: Objective: This study assessed the impact of systematic review and data extraction experience on the accuracy and efficiency of data extraction in systematic reviews.Study Design and Setting: We conducted a prospective cross-sectional study from October to December 2006. Participants were classified as having minimal, moderate, or substantial experience in systematic reviews and data extraction. Three studies on insomnia treatment were extracted. Our primary outcome was the accuracy of data extraction. Data sets of each experience level were analyzed for errors in data extraction and results of meta-analyses. Additionally, the time required for completion of data extraction was compared.Results: Error rates were similar across the various levels of experience and ranged from 28.3% to 31.2%. Mean rates for errors of omission (11.3–16.4%) were generally lower than those for errors of inaccuracy (13.9–17.9%). There were no significant differences in error rates or accuracy of meta-analysis results between groups. Time required approached significance, with minimally experienced participants requiring the most time.Conclusion: Overall, there were high error rates by participants at all experience levels; however, time required for extraction tended to decrease with experience. These results illustrate the need to develop strategies aimed at mastery of data extraction, rather than reliance on previous data extraction experience alone.Systematic review data extraction: cross-sectional study showed that experience did not increase accuracyJennifer Horton, Ben Vandermeer, Lisa Hartling, Lisa Tjosvold, Terry P. Klassen, Nina Buscemi10.1016/j.jclinepi.2009.04.007Journal of Clinical Epidemiology 63, 3 (2010)2009-08-17Journal of Clinical Epidemiology2009-08-17633S0895-4356(10)X0002-6Systematic Reviews and Meta Analysis289298http://www.jclinepi.com/article/PIIS0895435609002066/abstract?rss=yesAbstract: Objective: In cases of insufficiently controlled blood pressure, it is important for practitioners to distinguish between “nonadherence” and “nonresponse” to antihypertensive drug treatment. A reliable and valid adherence measurement based on the patient's self-report may be helpful in daily practice.Study Design and Setting: In a primary care sample with 353 hypertensive patients, we applied two self-rating instruments to assess medication adherence (the “Hill-Bone Compliance to High Blood Pressure Therapy Scale” and Morisky's “Self-Reported Measure of Medication Adherence”) and compared their psychometric properties.Results: Both scales showed low acceptability and insufficiency to moderate internal consistency (Cronbach's α=0.25 and 0.73, respectively). Their convergent validity as indexed by kappa=0.39 could be judged as “fair” at best. Testing the power to predict blood pressure >140/90mmHg, both scales showed an accuracy of 57% and 62%, respectively. The positive likelihood, that is, the increase in likelihood of high blood pressure in cases of nonadherence was 1.00 and 1.32, respectively.Conclusion: The use of both scales cannot be recommended. They showed considerable floor effects, and their ability to identify medication adherence was inconsistent for nearly every third patient. The power of both scales to predict uncontrolled blood pressure was essentially a chance. The underlying conceptual framework of medication adherence therefore needs to be rethought.Comparison of two self-rating instruments for medication adherence assessment in hypertension revealed insufficient psychometric propertiesJanka Koschack, Gabriella Marx, Jörg Schnakenberg, Michael M. Kochen, Wolfgang Himmel10.1016/j.jclinepi.2009.06.011Journal of Clinical Epidemiology 63, 3 (2010)2009-09-18Journal of Clinical Epidemiology2009-09-18633S0895-4356(10)X0002-6Original Articles299306http://www.jclinepi.com/article/PIIS0895435609001772/abstract?rss=yesAbstract: Objective: To assess the relationship between mortality and impairment and decline in a specific executive cognitive function, the capacity for behavioral self-regulation.Study Design and Setting: This study examined the association between mortality and baseline and 22-month decline in the capacity for behavioral self-regulation, as measured by the Behavioral Dyscontrol Scale, among 1,293 participants of the San Luis Valley Health and Aging Study (SLVHAS), a population-based longitudinal study. The Behavioral Dyscontrol Scale and a measure of overall mental status, the Mini-Mental State Examination, were administered at baseline and follow-up interviews. Cox regression was used to examine baseline and decline in capacity for behavioral self-regulation as possible predictors of mortality.Results: Baseline Behavioral Dyscontrol Scale score was predictive of mortality, independent of demographics and comorbidity count (hazard ratio [HR]=1.07; 95% confidence interval [CI]: 1.04, 1.09). It remained a significant predictor with further adjustment for Mini-Mental State Examination score. Decline in this specific executive cognitive function was associated with mortality after adjustment for covariates and baseline cognitive scores (HR=1.09; 95% CI: 1.04, 1.13).Conclusion: Thus, both baseline capacity for behavioral self-regulation and its decline over time predicted mortality in the SLVHAS cohort. These associations may partly be attributed to maintaining the ability for self-care. Understanding how specific forms of impairment contribute to mortality may help identify patients who could benefit from early intervention.Executive function (capacity for behavioral self-regulation) and decline predicted mortality in a longitudinal study in Southern ColoradoE. Amirian, Judith Baxter, Jim Grigsby, Douglas Curran-Everett, John E. Hokanson, Lucinda L. Bryant10.1016/j.jclinepi.2009.06.004Journal of Clinical Epidemiology 63, 3 (2010)2009-08-28Journal of Clinical Epidemiology2009-08-28633S0895-4356(10)X0002-6Original Articles307314http://www.jclinepi.com/article/PIIS089543560900184X/abstract?rss=yesAbstract: Objectives: To conduct the initial modern measurement theory analyses because of its many advantages on the Tampa Scale for Kinesiophobia and emerging evidence suggesting that fear of movement influences functional disability in people with arthritis.Study Design and Setting: Secondary analysis of 347 participants from a randomized controlled trial evaluating The People with Arthritis Can Exercise program. The original Tampa Scale for Kinesiophobia has 17 items and we collected 16 items (excluding item 6). An item response theory analysis was conducted using the graded response model in MULTILOG. Before this, a series of factor analyses assessed the unidimensionality assumption of this model.Results: Based on the factor analyses, we removed the reverse-coded items (4, 8, 12, and 16). The item response theory analysis revealed that item 13 had an exceedingly low slope and was dropped.Conclusion: Item response theory analyses looked at each item's performance and we can strongly suggest using our modified scale (11 items out of the 16 items), which provides relatively uniform precision of measurement across a wide range of fear of movement in people with arthritis. The item parameters from this study can build a computerized adaptive testing for this scale.First item response theory analysis on Tampa Scale for Kinesiophobia (fear of movement) in arthritisThelma J. Mielenz, Michael C. Edwards, Leigh F. Callahan10.1016/j.jclinepi.2009.04.011Journal of Clinical Epidemiology 63, 3 (2010)2009-09-07Journal of Clinical Epidemiology2009-09-07633S0895-4356(10)X0002-6Original Articles315320http://www.jclinepi.com/article/PIIS0895435609002078/abstract?rss=yesAbstract: Objective: To investigate how a multilevel item response theory (IRT) model for longitudinal dependent data could provide average and individual quality-of-life outcomes of low-vision rehabilitation.Study Design and Setting: In a nonrandomized longitudinal design, visually impaired older patients (n=296) were referred to multidisciplinary rehabilitation or to an optometric service. The five-dimensional Low Vision Quality of Life Questionnaire was administered at four time points. The IRT model was characterized by the graded response model for rating scales. Covariates were added to the model, mainly to correct for missing data. The invariance assumption across time points was investigated.Results: Average and individual rehabilitation effects were estimated. For multidisciplinary rehabilitation, significant average deterioration was seen on three dimensions after 4.4 years. Many individuals in the optometric service group significantly improved on the “reading small print” dimension (18.5%); in both groups, many individuals significantly deteriorated on “visual (motor) skills” (22.2–30.0%). Invariance across time points could be assumed for all dimensions, except for “adjustment.” Gender, education, visual acuity, and health status were significantly associated with the outcome.Conclusion: We present how a multilevel IRT model can be applied to describe longitudinal dependent vision-related quality-of-life data, while focusing on average and individual effects.A multilevel item response theory model was investigated for longitudinal vision-related quality-of-life dataRuth M.A. van Nispen, Dirk L. Knol, Han J. Neve, Ger H.M.B. van Rens10.1016/j.jclinepi.2009.06.012Journal of Clinical Epidemiology 63, 3 (2010)2009-09-22Journal of Clinical Epidemiology2009-09-22633S0895-4356(10)X0002-6Original Articles321330http://www.jclinepi.com/article/PIIS0895435609002200/abstract?rss=yesAbstract: Objectives: Compare three commonly used methods to combine the impacts of multiple health conditions on SF-6D health utility scores.Study Design and Setting: We used data from the 1998–2004 Medicare Health Outcomes Survey to compare three commonly suggested models of multiple health conditions' impacts on health-related quality of life: additive, minimum, and multiplicative. We modeled SF-6D scores using information about 15 health conditions, both unadjusted and adjusted for age, sex, education, and income. Model performance was assessed using mean squared error, mean predictive error by number of health conditions, and mean predictive error for groups with specific combinations of health conditions.Results: Ninety-five thousand one hundred ninety-five observations were used for model estimation, and 94,794 observations were used for model testing. The adjusted models always had better performance than the unadjusted models. The multiplicative model showed smaller mean predictive error than the other models in both those younger than 65 years and those 65 years and older. Mean predictive error for the multiplicative model was generally within the minimally important difference of the SF-6D.Conclusion: All tested models are imperfect in these Medicare data, but the multiplicative model performed best.Three methods tested to model SF-6D health utilities for health states involving comorbidity/co-occurring conditionsJanel Hanmer, David Vanness, Ronald Gangnon, Mari Palta, Dennis G. Fryback10.1016/j.jclinepi.2009.06.013Journal of Clinical Epidemiology 63, 3 (2010)2009-11-09Journal of Clinical Epidemiology2009-11-09633S0895-4356(10)X0002-6Original Articles331341http://www.jclinepi.com/article/PIIS0895435609003023/abstract?rss=yesOn the validity of meta-analyses: exhaustivity must be warranted, exclusion of duplicate patients tooElodie Pambrun, Vincent Bouteloup, Rodolphe Thiébaut, Julien Asselineau, Victor de Lédinghen, Paul Perez, Steering Committee of the Transient Elastography Individual Patient Data meta-analysis Study (TE IPD Study)10.1016/j.jclinepi.2009.07.021Journal of Clinical Epidemiology 63, 3 (2010)2010-01-18Journal of Clinical Epidemiology2010-01-18633S0895-4356(10)X0002-6Letters to the Editor342343http://www.jclinepi.com/article/PIIS0895435609003035/abstract?rss=yesIn their communication, Pambrun et al. comment on the well-known problems of literature-based meta-analysis in general and, especially, on the problem of including duplicate information from different publications in a recent literature-based meta-analysis on the diagnostic accuracy of transient elastography for the staging of liver fibrosis .Response to letter: On the validity of meta-analysisMireen Friedrich-Rust, Eva Herrmann10.1016/j.jclinepi.2009.09.008Journal of Clinical Epidemiology 63, 3 (2010)2010-01-08Journal of Clinical Epidemiology2010-01-08633S0895-4356(10)X0002-6Letters to the Editor343343