BackgroundAs statisticians develop new methodological approaches, there are many factors that influence whether others will utilize their work. This paper is a bibliometric study that identifies and quantifies associations between characteristics of new biostatistics methods and their citation counts. Of primary interest was the association between numbers of citations and whether software code was available to the reader.MethodsStatistics journal articles published in 2010 from 35 statistical journals were reviewed by two biostatisticians. Generalized linear mixed models were used to determine which characteristics (author, article, and journal) were independently associated with citation counts (as of April 1, 2017) in other peer-reviewed articles.ResultsOf 722 articles reviewed, 428 were classified as new biostatistics methods. In a multivariable model, for articles that were not freely accessible on the journal’s website, having code available appeared to offer no boost to the number of citations (adjusted rate ratio = 0.96, 95% CI = 0.74 to 1.24, p = 0.74); however, for articles that were freely accessible on the journal’s website, having code available was associated with a 2-fold increase in the number of citations (adjusted rate ratio = 2.01, 95% CI = 1.30 to 3.10, p = 0.002). Higher citation rates were also associated with higher numbers of references, longer articles, SCImago Journal Rank indicator (SJR), and total numbers of publications among authors, with the strongest impact on citation rates coming from SJR (rate ratio = 1.21 for a 1-unit increase in SJR; 95% CI = 1.11 to 1.32).ConclusionThese analyses shed new insight into factors associated with citation rates of articles on new biostatistical methods. Making computer code available to readers is a goal worth striving for that may enhance biostatistics knowledge translation.

BackgroundIn medical practice, clinically unexpected measurements might be quite properly handled by the remeasurement, removal, or reclassification of patients. If these habits are not prevented during clinical research, how much of each is needed to sway an entire study?Methods and ResultsBelieving there is a difference between groups, a well-intentioned clinician researcher addresses unexpected values. We tested how much removal, remeasurement, or reclassification of patients would be needed in most cases to turn an otherwise-neutral study positive. Remeasurement of 19 patients out of 200 per group was required to make most studies positive. Removal was more powerful: just 9 out of 200 was enough. Reclassification was most powerful, with 5 out of 200 enough. The larger the study, the smaller the proportion of patients needing to be manipulated to make the study positive: the percentages needed to be remeasured, removed, or reclassified fell from 45%, 20%, and 10% respectively for a 20 patient-per-group study, to 4%, 2%, and 1% for an 800 patient-per-group study. Dot-plots, but not bar-charts, make the perhaps-inadvertent manipulations visible. Detection is possible using statistical methods such as the Tadpole test.ConclusionsBehaviours necessary for clinical practice are destructive to clinical research. Even small amounts of selective remeasurement, removal, or reclassification can produce false positive results. Size matters: larger studies are proportionately more vulnerable. If observational studies permit selective unblinded enrolment, malleable classification, or selective remeasurement, then results are not credible. Clinical research is very vulnerable to “remeasurement, removal, and reclassification”, the 3 evil R's.

A dataset that is useful for training in applied biostatistics, in the context of biomedical research

Background and PurposeSpontaneous intracerebral hemorrhage (ICH) is a devastating form of stroke with a poor prognosis overall. We conducted a systematic review and meta-analysis to identify and describe factors associated with early neurologic deterioration (END) after ICH.MethodsWe sought to identify any factor which could be prognostic in the absence of an intervention. The Cochrane Library, EMBASE, the Global Health Library, and PubMed were searched for primary studies from the years 1966 to 2012 with no restrictions on language or study design. Studies of patients who received a surgical intervention or specific experimental therapies were excluded. END was defined as death, or worsening on a reliable outcome scale within seven days after onset.Results7,172 abstracts were reviewed, 1,579 full-text papers were obtained and screened. 14 studies were identified; including 2088 patients. Indices of ICH severity such as ICH volume (univariate combined OR per ml:1.37, 95%CI: 1.12–1.68), presence of intraventricular hemorrhage (2.95, 95%CI: 1.57–5.55), glucose concentration (per mmol/l: 2.14, 95%CI: 1.03–4.47), fibrinogen concentration (per g/l: 1.83, 95%CI: 1.03–3.25), and d-dimer concentration at hospital admission (per mg/l: 4.19, 95%CI: 1.88–9.34) were significantly associated with END after random-effects analyses. Whereas commonly described risk factors for ICH progression such as blood pressure, history of hypertension, and ICH growth were not.ConclusionsThis study summarizes the evidence to date on early ICH prognosis and highlights that the amount and distribution of the initial bleed at hospital admission may be the most important factors to consider when predicting early clinical outcomes.

Data from previous studies examining patient-related risk factors for aseptic loosening in patients who have undergone hip and knee arthroplasty.

*It is noted that multiple studies reported on more than a single cancer site and the total does not therefore equal 34.1A study reported increased risk for males but no increase for females.2A study reported increased risk for total population, but no increase when examined by gender.3A study reported increased risk for males but no increase for females.

Number of patients that can be manipulated before, on average, the manipulation is detected by the “tadpole test”.

This document provides technical details on boosted beta regression, as well as the full list of predictor variables used for the analysis of the NLA data. (PDF)

a, b : Scheffé test (the means with the same letter were significantly different).

HR: Hazard ratio; PCS: prospective cohort study; RCS: retrospective cohort study; RCT: randomized controlled trial; RR: relative risk; NA: not available.

: X2 test for the variables; THA: total hip arthroplasty; TKA: total knee arthroplasty.

Observed factor level counts (for which dvt = 1) for binary risk factors in each study of the DVT case study. Entries are left blank for studies that did not measure the corresponding factor.

Classification of statistical methods as reported in journals.

Rate Ratios (RRs) for cancers analysed for British White & British Indian children in Leicester.

PRISMA checklist. (DOC)

OR, odds ratio; CI, confidence interval; NA, not available; Large, cases ≥100; Small, cases

aALLRT participants 59 years were excluded due to limited sample size.

Notes:– All examples originated from the literature regarding portal vein thrombosis.– In every example, the same study was published in two different journals.– All literatures were expressed in Vancouver reference type.– Bold and italics formatting indicated the different styles between index and redundant paper(s).

File S1 includes Appendix S1, Appendix S2, Appendix S3, Appendix S4. Appendix S1: Search terms used to identify studies of one year mortality on antiretroviral therapy. Appendix S2: Full citations for studies reviewed. Appendix S3: Illustration of a distribution used to impute CD4 count with bands. Appendix S4: CD4 coefficient (bottom) and model fit (F-statistic – top) for the relationship between one year mortality on ART and baseline CD4 count using varying assumptions about the amount of mortality among those lost to follow-up. (DOCX)

BMI: body mass index; THA: total hip arthroplasty; TKA: total knee arthroplasty.Adjusted for age, sex, BMI, place of residence, diagnosis, and all comorbid conditions.