Systematic Selection Bias Research Articles

Influence of School Type and Class Level on Mean Caries Experience in 12-Year-Olds in Serial Cross-Sectional National Oral Health Survey in Germany-Proposal to Adjust for Selection Bias.

Caries data from the 2016 NOHSC were adjusted for each federal state on the basis of two additional large-scale datasets for school type and class level. Twelve-year-olds in all grades in Saxony-Anhalt (n = 96,842) exhibited significantly higher DMFT values than 12-year-olds in 6th grade (n = 76,456; +0.10 DMFT; ~14.2%, p < 0.001). Adjustments for school type had effects on DMFT on the level of federal states but almost balanced out on the national level (-0.01 DMFT; ~2%). Due to putatively similar structures of the federal states, the national mean DMFT for 12-year-olds in the latest NOHSC (2016; n = 55,002) was adjusted from 0.44 to 0.50 DMFT, correcting for selection bias. Selection bias in this NOHSC leads to an underestimation of caries levels by about 15%. Due to very low caries experience in children in Germany, these precise adjustments (+0.06 DMFT) have only a minor effect on interpretations of the national epidemiologic situation. Consequently, other national caries studies worldwide using the robust marker of DMFT should also adjust for systematic selection bias related to socio-economic background rather than increasing efforts in examination strategy.

Selection bias in multi-technology auctions: How to quantify and assess efficiency implications in renewable energy auctions

We develop a concept to identify and quantitatively assess technology selection bias in multi-technology renewable energy auctions. We show that simple price rules are insufficient to efficiently select winning projects in multi-technology tenders when they incorporate individual costs of producers only, and exclude system effects, market benefits and external costs. With our concept of quantifying unit social value, all relevant external elements can be incorporated in the evaluation. The introduction of an objective measure (as difference of differences between unit social value and auction bid prices) allows the quantitative assessment of systematic selection bias between technologies under different remuneration designs. We illustrate our concept by applying it to generic European renewable energy technologies (wind and solar energy) and major applied remuneration types (contract-for-difference and fixed premium schemes). We also show that are concept is applicable for real auction data, by presenting a case study about Italy. The main conclusion of the study is that selection by price only, can constitute a systematic bias for all investigated remuneration designs. Both schemes are in our case biased toward a given technology. The internalisation of external cost may not necessarily lead to better social outcomes in the selection of auction winners, as it can be overshadowed by the initial bias of the chosen remuneration design. We conclude that considerate design, including the potential differentiation of rules, e.g., through introduction of a technology bonus, to minimise selection bias, and careful monitoring of the competitive situation between participating technologies in a multi-technology auction are required to ensure efficient selection.

Credible Mendelian Randomization Studies in the Presence of Selection Bias Using Control Exposures.

Selection bias is increasingly acknowledged as a limitation of Mendelian randomization (MR). However, few methods exist to assess this issue. We focus on two plausible causal structures relevant to MR studies and illustrate the data-generating process underlying selection bias via simulation studies. We conceptualize the use of control exposures to validate MR estimates derived from selected samples by detecting potential selection bias and reproducing the exposure–outcome association of primary interest based on subject matter knowledge. We discuss the criteria for choosing the control exposures. We apply the proposal in an MR study investigating the potential effect of higher transferrin with stroke (including ischemic and cardioembolic stroke) using transferrin saturation and iron status as control exposures. Theoretically, selection bias affects associations of genetic instruments with the outcome in selected samples, violating the exclusion-restriction assumption and distorting MR estimates. Our applied example showing inconsistent effects of genetically predicted higher transferrin and higher transferrin saturation on stroke suggests the potential selection bias. Furthermore, the expected associations of genetically predicted higher iron status on stroke and longevity indicate no systematic selection bias. The routine use of control exposures in MR studies provides a valuable tool to validate estimated causal effects. Like the applied example, an antagonist, decoy, or exposure with similar biological activity as the exposure of primary interest, which has the same potential selection bias sources as the exposure–outcome association, is suggested as the control exposure. An additional or a validated control exposure with a well-established association with the outcome is also recommended to explore possible systematic selection bias.

ASPECTS OF THE USE OF PROPENSITY SCORE MATCHING METHODS IN NEUROSURGERY

ABSTRACT Objective Observational studies and register data provide researchers with ample opportunities to obtain answers to questions that randomized controlled trials cannot answer for organizational or ethical reasons. One of the most common tools for solving this problem is the use of propensity score matching (PSM) methods. The purposes of our study were to compare various models and algorithms for selecting PSM parameters, using retrospective clinical data, and to compare the results obtained using the PSM method with those of prospective studies. Methods The results of two studies (randomized prospective and retrospective) conducted at the Novosibirsk Research Institute of Traumatology and Orthopedics were used for comparative analysis. The trials aimed to study the effectiveness and safety of surgical treatment of degenerative dystrophic lesions in the lumbar spine. We compared the results using the recommended PSM parameters (caliper=0.2 and 0.6) the propensity score is the probability of assignment to one treatment conditional on a subject’s measured baseline covariates. Propensity-score matching is increasingly being used to estimate the effects of exposures using observational data. In the most common implementation of propensity-score matching, pairs of treated and untreated subjects are formed whose propensity scores differ by at most a pre-specified amount (the caliper widthand the caliper values often used in real-life studies (0.05, 0.1, 0.25, 0.5, and 0.8) with the those obtained in a similar prospective study. Results After eliminating systematic selection bias, the results of the retrospective and randomized prospective studies were qualitatively comparable. Conclusion The results of this study provide recommendations for the use of PSM: when evaluating efficacy scores in neurosurgical studies (with a sample size &lt; 150 patients), we recommend matching on the logit of the propensity score using calipers of width equal to 0.6 of the standard deviation of the logit of the propensity score. Level of evidence V; Type of study is expert opinion.

A New Approach to Estimating the Relation between Audit Fees and Financial Misconduct

SYNOPSISResearch on the association between abnormal audit fees (measuring audit effort) and financial misconduct has produced mixed results. The use of actual misstatements in this research creates small-sample inferences, introduces systematic selection bias, and reduces the scope of sample coverage. In this study we use a metric based on Benford's Law to analyze the impact of abnormal audit fees on the likelihood of misconduct. This measure is parsimonious, avoids selection bias, and can be computed for a large sample of public firms. Consistent with theory, we find that greater audit effort reduces the likelihood of misconduct and auditor resignations are more likely for clients with higher misconduct likelihood. Our findings are not driven by audit firm size, client size, the governance structure of the client, or economic bonding explanations. The effect is not subsumed when controlling for alternative misconduct measurement metrics and is robust across multiple tests to address endogeneity.JEL Classifications: G32; M41.

Current Issues in Randomized Clinical Trials of Neurodegenerative Disorders at Enrolment and Reporting: Diagnosis, Recruitment, Representativeness of Patients, Ethnicity, and Quality of Reporting.

The investigator is faced with several challenges when planning a randomized clinical trial (RCT). In the early phase, issues are particularly challenging for RCTs in neurodegenerative disorders (NDD). At the time of inclusion in the study, an early and accurate diagnosis is mandatory. Variability of diagnostic criteria, mostly based on clinical grounds, lag time between onset and enrolment, and phenotypic heterogeneity are the main drivers of diagnostic complexity. High-quality data in terms of diagnostic reliability, phenotypic description, follow-up, and evaluation of outcomes are key determinants and are highly conditioned by the expertise of the investigators and center recruitment rate. Representativeness of NDD patients is mandatory to postulate the generalizability of the results of RCTs. There is, however, a systematic selection bias in terms of age (more likely to be younger), sex (more likely to be male), ethnicity (more likely to be of European/Caucasian origin), and other prognostic factors (more likely to be favorable). In the publication phase, researchers need to report properly all of the main features of the RCT. Consolidated Standards of Reporting Trials (CONSORT) facilitates the report and interpretation of RCTs, but adherence to these guidelines needs to be improved. Several issues discussed in this review may alter the internal and external validity of an RCT. To date, the impact on phenotype at study entry has often been overlooked. A differential effect of the selection of subjects and of specific clinical and nonclinical features needs to be systematically explored in the RCT planning phase.

Sample selection may bias the outcome of an adolescent mental health survey: results from a five-year follow-up of 4171 adolescents

ObjectivesThe representativeness of the data is one of the main issues in evaluating the significance of research findings. Dropping out is common in adolescent mental health research, and may distort the results. Nevertheless, very little is known about the types of systematic bias that may affect studies in a) the informed consent phase and b) later in follow-up phases. Study designThe authors addressed this gap in knowledge in a five-year follow-up study on a sample of adolescents aged 13–18 years. MethodsThe data were collected using self-report questionnaires. The baseline sample consisted of 4171 adolescents, 1827 (43.8%) of whom gave consent to be contacted for a follow-up survey, but only 797 (19.1%) participated in the follow-up. Binary logistic regression models were used to explain the participation. ResultsYoung age, female gender, a high number of hobbies, good performance at school in the native language and general subjects, family disintegration such as divorce, high parental employment, and symptoms of depression and anxiety were associated with both consent and participation. However, the effect of mental health aspects was smaller than the effect of age and gender. ConclusionsThis study confirmed the possibility of systematic selection bias by adolescents' sociodemographic characteristics. The representativeness of the study sample might have been improved by more intense recruitment strategies.

Editorial Commentary: Accounting for Pertussis

The resurgence of measles and pertussis in the United States [1, 2] is a reminder of the importance of maximum immunization coverage, even in a developed country with relatively high rates of childhood immunization. The exact parameters of herd immunity may not be established for some vaccine-preventable diseases, but maximal effective use of available vaccines providespopulation-level immunity toprevent or reduce transmission. Evidencebased strategies published by the Advisory Committee on Immunization Practices (ACIP) for improved control of pertussis promote effective use [3, 4]. Observation of vaccine effectiveness in routine clinical use provides insight that may not be available from clinical efficacy trials and may be used to inform recommendations. The article by Witt et al [5] in this issue of Clinical Infectious Diseases presents a retrospective study of acellular pertussis vaccine effectiveness made possible by the unique circumstances of a community-wide outbreak of Bordetella pertussis infection affecting a large population under the care of an integrated healthcare system.This allowedestimation of vaccine effectiveness by age group. The relative duration of protection achieved with diphtheria toxoid–tetanus toxoid–acellular pertussis (DTaP) vaccines (introduced in theUnited States in the 1990s) remains a question [6], one made more relevant by the possibilities arising from the availability of pertussis antigen containing vaccines (tetanus toxoid– reduced diphtheria toxoid–acellular pertussis vaccine for adolescents and adults [Tdap]) that allow for booster dosing of individuals >7 years of age. The authors were able to demonstrate less than expected effectiveness of DTaP in 8to 12year-olds, while also demonstrating substantial overall effectiveness of pertussis vaccines. On the basis of their observations, they suggest evaluation of earlier and more frequent doses of pertussis containing vaccine (Tdap) in outbreak settings, and even on a routine basis. More studies of strategic use of Tdap as a booster dose vaccine would be worthwhile, especially after the broad experience with the vaccine and its safety profile. The ACIP already recommends Tdap use in undervaccinated children as young as 7 years of age [3]. A question might be raised as to the use of polymerase chain reaction (PCR) as the exclusive diagnostic test for pertussis in the Witt et al study, especially as the minimal cough duration in the practice guidance was only 1 week. PCR is a widely used, sensitive test that is faster and easier to perform than culture and can identify evidence of B. pertussis infection in situations where culture results are negative. However, high sensitivity and low specificity, with pseudo-outbreaks described due to false positives and environmental contamination with B. pertussis genome [7, 8], suggest that PCR only be used with care and for patients with cough duration of≥2 weeks and a clinical picture consistent with pertussis. However, the observations by Witt et al [5] occurred in the presence of a widespread, welldocumented stateand community-wide outbreak of pertussis (high prevalence of diseaseimprovingpositivepredictivevalue), and testing was done in a central laboratory, using a closed system test platform, and simultaneous testing for Bordetella parapertussis. Furthermore, rates of testing were high across the spectrum of ages, but rates of positive PCR results are very different by age, so false-positive test results do not seem to be a factor contributing to the results unless there was a systematic bias in selection for testing by age. Pertussis remains a threat to infants, in whom protection awaits the initial doses of the DTaP vaccine series. Public health recommendations for the use of Tdap vaccines have been a work in progress since licensure in the United States, and approaches have been designed to protect older children, adolescents, and adults, in large part to protect infants they may expose to pertussis (as symptomatic and asymptomatic cases to the extent that overall reduction in incidence may reduce asymptomatic carriage). Initially, Tdap Received 27 February 2011; accepted 1 March 2012; electronically published 15 March 2012. Correspondence: Altred De Maria Jr, Bureau of Infectious Disease, Massachusetts Department of Public Health, William A. Hinton State Laboratory Institute, 305 South St, Jamaica Plain, MA 02130 (Alfred.DeMaria@state.ma.us). Clinical Infectious Diseases 2012;54(12):1736–8 © The Author 2012. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: journals.permissions@ oup.com. DOI: 10.1093/cid/cis313

Talent Identification in Soccer: The Role of Maturity Status on Physical, Physiological and Technical Characteristics

The purpose of this paper is to provide a review around talent identification in soccer using physiological and technical testing procedures, and to summarise the issues associated with this process. The current research in soccer talent identification, among other sports, demonstrates a systematic bias in selection towards players born early in the year (i.e., relative age effect) and early maturers. From the studies investigating the physiological (e.g., power) and technical (e.g., dribbling) characteristics of players of different maturity status, early maturers had the tendency to perform better in these tests and therefore were likely to be more influential on the game and be recognised as more talented. When considering the current level of play and future success, elite youth and future professional players scored better in physiological and technical testing than recreational youth and future non-professional players, independently of maturity status. However, these testing procedures were not sensitive enough to distinguish youth elite from sub-elite or future national team from professional club players. Collectively, these studies demonstrated the need to use estimates of maturity status and subsequent appropriate analysis of data obtained from physiological and technical testing. When maturity is taken into account, these testing procedures can provide an indication of responsiveness to training load in youth players and an evaluation of potential to become a successful soccer player. However, these testing procedures should not be used as a marker of selection before full maturity is attainted and should be part of a multidimensional approach of talent identification considering the importance of other facets of the game at the highest level (e.g., perceptive-cognitive skills).

Sample bias from different recruitment strategies in a randomised controlled trial for alcohol dependence

Participants may be recruited from diverse sources for randomised controlled trials (RCT) of treatments for alcohol dependence. A mixed recruitment strategy might facilitate recruitment and increase generalisability at the expense of introducing systematic selection bias. The current study aims to compare the effects of recruitment method on socio-demographics, baseline illness characteristics, treatment retention and treatment outcome measures. A secondary analysis from a previous 12 week RCT of naltrexone, acamprosate and placebo for alcohol dependence was conducted. Participants (n = 169) were obtained via four channels of recruitment including in-patient and outpatient referral, live media and print media solicitation. Baseline parameters, retention in treatment and treatment outcomes were compared in these groups. Relative to in-patient subjects, those recruited via live and print media had significantly lower scores on taking steps, less in-patient rehabilitation admissions and less previous abstinence before entering the trial. Subjects recruited via print media had significantly lower scores of alcohol dependence relative to all other modes recruitment. There were no differences between recruitment strategies on treatment retention or compliance. At outcome, no significant effect of recruitment method was detected. These results suggest that different recruitment methods may be sourcing subjects with different baseline characteristics of illness. Nonetheless, these differences did not significantly impact on treatment retention or outcome, suggesting that in this population it was appropriate to recruit subjects from mixed sources.

The Part-Time Wage Gap in Norway: How Large is It Really?

Norwegian working-life legislation has strict rules against discrimination between full-time and part-time work. Partly as a consequence of this, a large proportion of Norwegian women work part-time. The purpose of this paper is to establish whether there are systematic differences between part-time and full-time workers regarding the selection process and earnings capacity. We find wage differences between part-time and full-time workers are small in Norway, and there is no evidence of systematic selection bias. The results may be explained by some features of the Norwegian labour market, including, equal rights for part-time workers, strict rules against the discrimination of part-time workers, and a generous family policy enabling women to combine work and family life.