Bias in meta-analytic research

doi:10.1016/0895-4356(92)90072-U

Journal of Clinical Epidemiology

Volume 45, Issue 8, August 1992, Pages 885-892

https://doi.org/10.1016/0895-4356(92)90072-U Get rights and content

Abstract

With the proliferation of meta-analyses in the medical literature have come conflicting studies. In addition, observance of guidelines for the performance of meta-analyses has been spotty. Bias may explain conflicting studies and differentiate carefully performed meta-analyses from others. Meta-analysts may fail to anticipate biases which threaten their study's validity. The three stages at which bias can be injected into a meta-analysis are finding studies, selection of the identified studies for the meta-analysis and extraction of data from the selected studies. This manuscript reviews specific types of bias which are common at each of these stages.

References (32)

M. Jenicek
Meta-analysis in medicine: where we are and where we want to go
J Clin Epidemiol
(1989)
P.J. Easterbrook et al.
Publication bias in clinical research
Lancet
(1991)
R. Rosenthal
The “file-drawer problem” and tolerance for null results
Psychol Bull
(1979)
P. Hewitt et al.
Perusing the literature: Methods of accessing MEDLINE and related databases
Controlled Ctin Trials
(1985)
K. Dickersin et al.
Perusing the literature: comparison of MEDLINE searching with a perinatal trials database
Controlled Clin Trials
(1985)
T. Poynard et al.
The retrieval of randomized clinical trials in liver disease from the medical literature: A comparison of MEDLARS and manual methods
Controlled Clin Trip1s
(1985)
F. Bernstein
The retrieval of randomized clinical trials in liver diseases from the medical literature: manual versus MEDLARS searches
Controlled Clin Trials
(1988)
M.J. Stampfer et al.
Effect of intravenous streptokinase on acute myocardial infarction: pooled results from randomized trials
N Engl J Med
(1982)
Effects of adjuvant tamoxifen and of cytotoxic therapy on mortality in early breast cancer: an overview of 61 randomized trials among 28,896 women
N Engl J Med
(1988)
N.H. Himel et al.
Adjuvant chemotherapy for breast cancer: a pooled estimate based on published randomized control trials
JAMA
(1986)

D.T. Felson et al.

Evidence for the superiority of immunosuppressive drugs and prednisone over prednisone alone in lupus nephritis: results of a pooled analysis

N Engl J Med

(1984)

H.S. Sacks et al.

Meta-analyses of randomized controlled trials

N Engl J Med

(1987)

J.M. Last

G.V. Glass et al.

Meta-analysis in Social Research

(1987)

R.J. Simes

Publication bias: The case for an international registry of clinical trials

J Clin Oncol

(1986)

C.B. Begg et al.

Publication bias and dissemination of clinical research

J Natl Cancer Inst

(1989)

Cited by (186)

Shelf life estimation of refrigerated vacuum packed beef accounting for uncertainty
2023, International Journal of Food Microbiology
This study estimates the shelf life of vacuum packed beef meat (three muscles: striploin (longissimus thoracis et lumborum, LTL), tenderloin (psoas major, PM) and outside chuck (trapezius thoracis, TT)) at refrigeration temperatures (0 °C–10 °C) based on modelling the growth of two relevant groups of spoilage microorganisms: lactic acid bacteria (LAB) and Enterobacteriaceae. The growth models were developed combining a two-step and a one-step approach. The primary modelling was used to identify the parameters affecting the growth kinetics, guiding the definition of secondary growth models. For LAB, the secondary model included the effect of temperature and initial pH on the specific growth rate. On the other hand, the model for Enterobacteriaceae incorporated the effect of temperature on the specific growth rate and the lag phase; as well as the effect of the initial pH on the specific growth rate, the lag phase and the initial microbial count. We did not observe any significant effect of the type of muscle on the growth kinetics. Once the equations were defined, the models were fitted to the complete dataset using a one-step approach. Model validation was carried out by cross-validation, mitigating the impact of an arbitrary division between training and validation sets. The models were used to estimate the shelf life of the product, based on the maximum admissible microbial concentration (7 log CFU/g for LAB, 5 log CFU/g for Enterobacteriaceae). Although LAB was the dominant microbiota, in several cases, both LAB and Enterobacteriaceae reached the critical concentration practically at the same time. Furthermore, in some scenarios, the end of shelf life would be determined by Enterobacteriaceae, pointing at the potential importance of non-dominant microorganisms for product spoilage. These results can aid in the implementation of effective control measures in the meat processing industry.
Turning awareness into behavior: Meta-analysis of household residential life energy transition behavior from the dual perspective of internal driving forces and external inducing forces
2023, Energy
The transition in household energy use from conventional to clean energy sources plays a critical role in reducing greenhouse gas emissions and mitigating climate change. In this study, we systematically considered the driving-inducing mechanisms of household residential energy transition from both external scenarios and internal psychological aspects. Based on data from 361,127 samples spanning 2005–2021, this study innovatively proposed the driving-inducing mechanism of the household energy transition. With the help of meta-analysis and meta-regression analysis (MRA), we comprehensively analyzed the psychological and behavioral differences in the process of household energy transition in different countries and different clean energy types, and revealed the evolution pattern of household residents from transition awareness to transition behavior in response to external scenarios and internal psychological changes. The results showed that the optimal drivers and inducers of household energy transition vary according to the diversity of clean energy types, and policy utility dislocation is the key external inducer that influences the transition from awareness to behavior of household residents. Further, the policy strategy should be tailored by region and type to better promote household energy transition.
The importance of what we cannot observe: Experimental limitations as a source of bias for meta-regression models in predictive microbiology
2023, International Journal of Food Microbiology
Meta-regression models have gained in popularity during the last years as a way to create more generic models for Microbial Risk Assessments that also include variability. However, as with most meta-analyses and empirical models, systematic biases in the data can result in inaccurate models. In this article, we define experimental bias as a type of selection bias due to the practical limitations of microbial inactivation experiments. Conditions with extremely high D-values (i.e. slow inactivation) need very long experimental runs to cause significant reductions. On the other hand, when the D-value is extremely low, not enough data points can be gathered before the microbial population is below the detection limit. Consequently, experimental designs favour conditions within a practical experimental range, introducing a selection bias in the D-values.
We demonstrate the impact of experimental bias in meta-regression models using numerical simulations. Models fitted to data with experimental bias overestimated the z-value and underestimated variability. We propose a rapid heuristic method to identify experimental bias in datasets, and we propose truncated regression to mitigate its impact in meta-regression models. Both methods were validated using simulated data. Thereafter the procedures were tested by building a meta-regression model for actual data for the inactivation of Bacillus cereus spores. We concluded that the dataset included experimental bias, and that it would cause an overestimation of the microbial resistance at high temperatures (>120 °C) for classical meta-regression models. This effect was mitigated when the model was built using truncated regression. In conclusion, we demonstrate that experimental bias could potentially result in inaccurate models for predictive microbiology. Therefore, checking for experimental bias should be a routine step in meta-regression modelling, and be included in guidelines on data analysis for meta-regression.
How can Blockchain technology support patent management? A systematic literature review
2023, International Journal of Information Management
Patent management tasks involve many stakeholders from diverse disciplines and hierarchies as well as from inside and outside the organization. By drawing on the transaction cost and stakeholder theories, Blockchain technology offers new opportunities to rethink the way in which stakeholders are organized and collaborate. Although Blockchain technology receives considerable attention from both the public and academia, the question of how Blockchain technology leverages patent management in particular has so far been neglected, although it holds great promise. Consequently, this is what we choose to be our basic research question. In response to it, we conduct a systematic literature review, following a procedure in three steps. Out of an initial sample of 714 articles, we select 52 as being particularly relevant, and classify them in accordance with the dimensions of the 7D Patent Management Maturity Model. Three main foci regarding the use of Blockchain technology emerge – i.e. for the generation, the enforcement, and the exploitation of patents. This study contributes to scholarship and management by showing that Blockchain technology mitigates environmental and behavioral uncertainty, while stimulating new governance forms and business models by reorganizing or discarding stakeholders.
Meta-analysis of growth performance indices of broiler chickens in response to turmeric (Curcuma longa L.) supplementation
2022, Animal Feed Science and Technology
Turmeric is currently used in different poultry species for their numerous beneficial effects. The existing body of knowledge has shown variable growth performance results in broiler chickens fed turmeric (Curcuma longa L.) supplemented diets. This investigation aimed to ascertain the effect of turmeric supplementation on growth performance of broiler chickens using meta-analysis. Four online databases were used to identify studies that investigated the impact of turmeric supplementation on feed intake (FI), feed conversion ratio (FCR) and average daily gain (ADG) in broiler chickens. One hundred and forty studies were retrieved from the search conducted on the four online databases and forty studies met the inclusion criteria for the meta-analysis. The random-effects model was used for data analysis. The effect size was calculated using a standardized mean difference (SMD) with 95% confidence intervals (CI). In comparison with the controls, turmeric supplementation significantly reduced FCR (SMD = - 0.44, 95% CI −0.57 to −0.32, I² = 88%), increased ADG (SMD = 0.44, 95% CI 0.30–0.57; I² = 89%) and decreased daily FI (SMD = −0.12, 95% CI −0.21 to −0.04 I² = 73%). Sub-analysis of dose-related response revealed statistically higher ADG (SMD = 0.48, 95% CI 0.33–0.62) in broiler chickens fed turmeric-supplemented diet at 0.2–11 g/kg feed. In contrast, turmeric-supplemented diet at 12 – 30 g/kg feed had no significant influence on ADG (SMD = 0.18, 95% CI −0.15 to 0.52). Significantly better FCR and ADG were recorded in broiler chickens fed turmeric-supplemented diet for 1–35 and 1–42 days, indicating that 1–35 and 1–42 days may be the optimal supplementation duration. However, there was no effect between feed intake and duration of turmeric supplementation. Birds from studies that fed turmeric-supplemented diets had improved growth performance compared to birds from studies that used turmeric extract. High heterogeneity was observed across the studies included in the meta-analysis. Meta-regression analysis found that studied covariates (supplementation level, duration of supplementation, turmeric presentation form and broiler strain) explained most of the sources of heterogeneity observed in the present meta-analysis. The mean effect size of 0.44 and 0.48 recorded for FCR and ADG, respectively suggesting that turmeric supplementation had a medium impact on growth performance indices in broiler chickens and could be used as a growth-enhancing agent. However, further research focusing on the optimal turmeric supplementation level and duration that optimize growth performance indices in broiler chickens is still needed.
A sustainable-development approach for self-adaptive cyber–physical system's life cycle: A systematic mapping study
2021, Journal of Systems and Software
Cyber–Physical Systems (CPS) refer to a new generation of systems where the cyber and physical layers are –strongly– interconnected. The development of these systems requires two fundamental parts. First, the design of sustainable architectures –centered on adaptation, throughout a System-Development Life-Cycle (SDLC)– to develop robust and economically profitable products. Second, the use of self-adaptive techniques to adjust CPSs to the evolving circumstances of their operation context. This work presents a systematic mapping study (SMS) that discusses different approaches used to develop self-adaptive CPSs (SA-CPSs) at each stage of the SDLC, focused on sustainability. The results show trends such as (i) Designs are not limited to particular application domains, (ii) Performance was the most commonly used attribute, and (iii) Monitor–Analyze–Plan–Execute over a shared Knowledge (MAPE-K) is the predominant feedback loop applied in the cyber layer. The results also raise challenges such as (i) How to design and evaluate sustainable SA-CPSs, (ii) How to apply unit and integration testing in the development of SA-CPSs, and (iii) How to develop feedback loops on SA-CPSs with the integration of machine-learning techniques.

View all citing articles on Scopus

View full text

Original articleBias in meta-analytic research

Abstract

J Clin Epidemiol

Lancet

Psychol Bull

Controlled Ctin Trials

Controlled Clin Trials

Controlled Clin Trip1s

Controlled Clin Trials

Effect of intravenous streptokinase on acute myocardial infarction: pooled results from randomized trials

N Engl J Med

Effects of adjuvant tamoxifen and of cytotoxic therapy on mortality in early breast cancer: an overview of 61 randomized trials among 28,896 women

N Engl J Med

Adjuvant chemotherapy for breast cancer: a pooled estimate based on published randomized control trials

JAMA

Evidence for the superiority of immunosuppressive drugs and prednisone over prednisone alone in lupus nephritis: results of a pooled analysis

N Engl J Med

Meta-analyses of randomized controlled trials

N Engl J Med

Meta-analysis in Social Research

Publication bias: The case for an international registry of clinical trials

J Clin Oncol

Publication bias and dissemination of clinical research

J Natl Cancer Inst

Original article
Bias in meta-analytic research