PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
 
BMJ. 2011; 342: d631.
Published online Mar 16, 2011. doi:  10.1136/bmj.d631
PMCID: PMC3175742
Strengthening the reporting of genetic risk prediction studies: the GRIPS statement
A Cecile J W Janssens, associate professor of translational epidemiology,corresponding author1 John P A Ioannidis, C F Renhborg professor in disease prevention,2,3,4,5,6 Cornelia M van Duijn, professor of genetic epidemiology,1 Julian Little, Canada research chair in human genome epidemiology ,7 and Muin J Khoury, director8, for the GRIPS Group
1Department of Epidemiology, Erasmus University Medical Center, Rotterdam, Netherlands
2Department of Hygiene and Epidemiology, University of Ioannina School of Medicine, Ioannina, Greece
3Biomedical Research Institute, Foundation for Research and Technology, Ioannina
4Department of Medicine, Tufts University School of Medicine, Boston, Massachusetts, USA
5Center for Genetic Epidemiology and Modeling and Tufts CTSI, Institute for Clinical Research and Health Policy Studies, Tufts Medical Center, Boston
6Stanford Prevention Research Center, Stanford University School of Medicine, Stanford, California, USA
7Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada
8Office of Public Health Genomics, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
corresponding authorCorresponding author.
Correspondence to: ACJW Janssens ; a.janssens/at/erasmusmc.nl
The recent successes of genome-wide association studies and the promises of whole genome sequencing fuel interest in the translation of this new wave of basic genetic knowledge to healthcare practice. Knowledge about genetic risk factors may be used to target diagnostic, preventive, and therapeutic interventions for complex disorders based on a person’s genetic risk or to complement existing risk models based on classic non-genetic factors such as the Framingham risk score for cardiovascular disease. Implementation of genetic risk prediction in healthcare requires a series of studies that encompass all phases of translational research,1 2 starting with a comprehensive evaluation of genetic risk prediction.
With increasing numbers of discovered genetic markers that can be used in future genetic risk prediction studies, it is crucial to enhance the quality of the reporting of these studies, since valid interpretation could be compromised by the lack of reporting of key information. Information that is often missing includes details in the description of how the study was designed and conducted (eg, how genetic variants were selected and coded, how risk models or genetic risk scores were constructed, and how risk categories were chosen), or how the results should be interpreted. An appropriate assessment of the study’s strengths and weaknesses is not possible without this information. There is ample evidence that prediction research often suffers from poor design and bias, and these may also have an impact on the results of the studies and on models of disease outcomes based on these studies.3 4 5 Although most prognostic studies published to date claim significant results,6 7 very few translate to clinically useful applications. Just as for observational epidemiological studies,8 poor reporting complicates the use of the specific study for research, clinical, or public health purposes and hampers the synthesis of evidence across studies.
Reporting guidelines have been published for various research designs,9 and these contain many items that are also relevant to genetic risk prediction studies. In particular, the guidelines for genetic association studies (STREGA) have relevant items on the assessment of genetic variants, and the guidelines for observational studies (STROBE) have relevant items about the reporting of study design. The guidelines for diagnostic studies (STARD) and those for tumour marker prognostic studies (REMARK) include relevant items about test evaluation; the REMARK guidelines also have relevant items about risk prediction.5 10 11 12 However, none of these guidelines are fully suited to genetic risk prediction studies, an emerging field of investigation with specific methodological issues that need to be addressed, such as the handling of large numbers of genetic variants (from 10s to 10 000s) and flexibility in handling such large numbers in analyses. We organised a two day workshop with an international group of risk prediction researchers, epidemiologists, geneticists, methodologists, statisticians, and journal editors to develop recommendations for the reporting of genetic risk prediction studies (GRIPS).
Genetic risk prediction studies typically develop or validate models that predict the risk of disease, but they are also being investigated for use in predicting prognostic outcome, treatment response, or treatment related harms. Risk prediction models are statistical algorithms, which may be simple genetic risk scores (for example, risk allele counts), may be based on regression analyses (for example, weighted risk scores or predicted risks), or may be based on more complex analytical approaches such as support vector machine learning or classification trees. The risk models may be based on genetic variants only or include both genetic and non-genetic risk factors.13
The 25 items of the GRIPS statement are intended to maximise the transparency, quality, and completeness of reporting on research methodology and findings in a particular study. It is important to emphasise that these recommendations are guidelines only for how to report research and do not prescribe how to perform genetic risk prediction studies. The guidelines do not support or oppose the choice of any particular study design or method—for example, the guidelines recommend that the study population should be described but do not specify which population is preferred in a particular study.
The intended audience for the reporting guidelines is broad and includes epidemiologists, geneticists, statisticians, clinician scientists, and laboratory based investigators who undertake genetic risk prediction studies, as well as journal editors and reviewers who have to appraise the design, conduct, and analysis of such studies. In addition, it includes “users” of such studies who wish to understand the basic premise, design, and limitations of genetic prediction studies in order to interpret the results for their potential application in healthcare. These guidelines are also intended to ensure that essential data from future genetic risk prediction studies are presented in standardised form, which will facilitate information synthesis as part of systematic reviews and meta-analyses.
Items presented in the checklist are relevant for a wide array of risk prediction studies, because GRIPS focuses on the main aspects of the design and analysis of risk prediction studies. GRIPS does not address randomised trials that may be performed to test risk models, nor does it specifically address decision analyses, cost effectiveness analyses, assessment of healthcare needs, or assessment of barriers to healthcare implementation.14 Once the performance of a risk model has been established, these next steps toward implementation require further evaluation.10 15 For the reporting of these studies, which go beyond the assessment of genetic risk models as such, additional requirements apply. However, proper documentation of genetic predictive research according to GRIPS might facilitate the translation of research findings into clinical and public health practice.
The GRIPS statement was developed by a multidisciplinary panel of 25 risk prediction researchers, epidemiologists, geneticists, methodologists, statisticians, and journal editors, seven of whom were also part of the STREGA initiative.11 They attended a two day meeting in Atlanta, Georgia, USA, in December 2009 that was sponsored by the US Centers for Disease Control and Prevention on behalf of the Human Genome Epidemiology Network (HuGENet).16 Participants discussed a draft version of the guidelines that was prepared and distributed before the meeting. This draft version was developed on the basis of existing reporting guidelines—namely, STREGA,11 REMARK,5 and STARD.12 These were selected out of all available guidelines (see www.equator-network.org) because of their focus on observational study designs and genetic factors (STREGA), prediction models (REMARK), and test evaluation (REMARK and STARD).
During the meeting, methodological issues pertinent to risk prediction studies were addressed in presentations. Workshop participants were asked to change, combine, or delete proposed items and add additional items if necessary. Participants had extensive post-meeting electronic correspondence. To harmonise our recommendations for genetic risk prediction studies with previous guidelines, we chose the same wording for the items wherever possible. Finally, we tried to create consistency with previous guidelines for the evaluation of risk prediction studies of cardiovascular diseases and cancer.2 17 The tabletable gives the final version of the checklist.
Table thumbnail
Reporting recommendations for evaluations of risk prediction models that include genetic variants
GRIPS explanation and elaboration article
To accompany this GRIPS statement, we have written an explanation and elaboration document,18 modelled on those developed for other reporting guidelines.19 20 21 22 The document illustrates each item with at least one published example that we consider transparent in reporting, explains the rationale for its inclusion in the checklist, and presents details of the items that need to be addressed to ensure transparent reporting. The explanation and elaboration document was produced after the meeting. The document was prepared by a small subgroup and shared with all workshop participants for additional revisions and final approval.
High quality reporting reveals the strengths and weaknesses of empirical studies, facilitates the interpretation of the scientific and healthcare relevance of the results—especially within the framework of systematic reviews and meta-analyses—and helps build a solid evidence base for moving genomic discoveries into applications in healthcare practice. The GRIPS guidelines were developed to improve the transparency, quality, and completeness of the reporting of genetic risk prediction studies. As outlined in the introduction, GRIPS does not prescribe how studies should be designed, conducted, or analysed, and therefore the guidelines should not be used to assess the quality of empirical studies.23 The guidelines should be used only to check whether all essential items are adequately reported.
Finally, the methodology for designing and assessing genetic risk prediction models is still developing. For example, newer measures of reclassification were first introduced in 2007,24 and several alternative reclassification measures have been proposed.25 Which measures to apply and when to use measures of reclassification are still subject to ongoing evaluation and discussion.26 Furthermore, alternative strategies for constructing risk models other than simple regression analyses are being explored, and these may add increased complexity to the reporting. In formulating the items of the GRIPS statement, these methodological advances were anticipated. It is for this reason that the GRIPS statement recommends how a study should be reported and not how a study should be conducted or analysed. Therefore, methodological and analytical developments will not immediately impact the validity and relevance of the items, but the GRIPS statement will be updated when this is warranted by essential new developments in the construction and evaluation of genetic risk models.
Summary points
  • The rapid and continuing progress in gene discovery for complex diseases is fuelling interest in the potential application of genetic risk models for clinical and public health practice
  • The number of studies assessing the predictive ability is steadily increasing, but the quality and completeness of reporting varies
  • A multidisciplinary workshop sponsored by the Human Genome Epidemiology Network developed a checklist of 25 items recommended for strengthening the reporting of genetic risk prediction studies (GRIPS), building on the principles established by prior reporting guidelines
  • These recommendations aim to enhance the transparency of study reporting and thereby improve the synthesis and application of information from multiple studies that might differ in design, conduct, or analysis
Notes
Members of the GRIPS group are: A Cecile J W Janssens, John P A Ioannidis, Sara Bedrosian, Paolo Boffetta, Siobhan M Dolan, Nicole Dowling, Isabel Fortier, Andrew N Freedman, Jeremy M Grimshaw, Jeffrey Gulcher, Marta Gwinn, Mark A Hlatky, Holly Janes, Peter Kraft, Stephanie Melillo, Christopher J O’Donnell, Michael J Pencina, David Ransohoff, Sheri D Schully, Daniela Seminara, Deborah M Winn, Caroline F Wright, Cornelia M van Duijn, Julian Little, and Muin J Khoury.
Funding: The workshop was sponsored by the Centers for Disease Control and Prevention on behalf of the Human Genome Epidemiology Network. Opinions in this paper do not necessarily represent the official position or policies of the authors’ employers, and the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: All authors have completed the unified competing interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare support from the Human Genome Epidemiology Network; ACWJ is supported by grants from the Erasmus University Medical Centre Rotterdam, the Centre for Medical Systems Biology in the framework of the Netherlands Genomics Initiative (NGI) and the VIDI grant of the Netherlands Organisation for Scientific Research (NWO) and Tufts CTSI is supported by the National Institutes of Health/National Center for Research Resources (UL1 RR025752); they have no other relationships or activities that could appear to have influenced the submitted work.
In order to encourage dissemination of the GRIPS statement, this article will also be published by Annals of Internal Medicine, BMJ, Circulation: Cardiovascular Genetics, European Journal of Clinical Investigation, European Journal of Epidemiology, European Journal of Human Genetics, Genetics in Medicine, Genome Medicine, and Journal of Clinical Epidemiology. A detailed explanation and elaboration document is published as supporting information.18
Provenance and peer review: Not commissioned; externally peer reviewed.
Notes
Cite this as: BMJ 2011;342:d631
1. Khoury MJ, Gwinn M, Yoon PW, Dowling N, Moore CA, Bradley L, et al. The continuum of translation research in genomic medicine: how can we accelerate the appropriate integration of human genome discoveries into health care and disease prevention? Genet Med 2007;9:665-74. [PubMed]
2. Hlatky MA, Greenland P, Arnett DK, Ballantyne CM, Criqui MH, Elkind MS, et al. Criteria for evaluation of novel markers of cardiovascular risk: a scientific statement from the American Heart Association. Circulation 2009;119:2408-16. [PMC free article] [PubMed]
3. Kyzas PA, Denaxa-Kyza D, Ioannidis JP. Quality of reporting of cancer prognostic marker studies: association with reported prognostic effect. J Natl Cancer Inst 2007;99:236-43. [PubMed]
4. Kyzas PA, Loizou KT, Ioannidis JP. Selective reporting biases in cancer prognostic factor studies. J Natl Cancer Inst 2005;97:1043-55. [PubMed]
5. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM, et al. Reporting recommendations for tumor marker prognostic studies (REMARK). Nat Clin Pract Oncol 2005;2:416-22. [PubMed]
6. Kyzas PA, Denaxa-Kyza D, Ioannidis JP. Almost all articles on cancer prognostic markers report statistically significant results. Eur J Cancer 2007;43:2559-79. [PubMed]
7. Tzoulaki I, Liberopoulos G, Ioannidis JP. Assessment of claims of improved prediction beyond the Framingham risk score. JAMA 2009;302:2345-2352. [PubMed]
8. Von Elm E, Egger M. The scandal of poor epidemiological research. BMJ 2004;329:868-9. [PMC free article] [PubMed]
9. Simera I, Moher D, Hoey J, Schulz KF, Altman DG. A catalogue of reporting guidelines for health research. Eur J Clin Invest 2010;40:35-53. [PubMed]
10. Von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med 2007;4:e296. [PMC free article] [PubMed]
11. Little J, Higgins JP, Ioannidis JP, Moher D, Gagnon F, von Elm E, et al. Strengthening the reporting of genetic association studies (STREGA): an extension of the STROBE statement. PLoS Med 2009;6:e22. [PMC free article] [PubMed]
12. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. BMJ 2003;326:41-4. [PMC free article] [PubMed]
13. Janssens ACJW, van Duijn CM. Genome-based prediction of common diseases: methodological considerations for future research. Genome Med 2009;1:20. [PMC free article] [PubMed]
14. Khoury MJ, Gwinn M, Ioannidis JP. The emergence of translational epidemiology: from scientific discovery to population health impact. Am J Epidemiol 2010;172:517-24. [PMC free article] [PubMed]
15. Moons KG, Altman DG, Vergouwe Y, Royston P. Prognosis and prognostic research: application and impact of prognostic models in clinical practice. BMJ 2009;338:b606. [PubMed]
16. Khoury MJ, Dorman JS. The Human Genome Epidemiology Network. Am J Epidemiol 1998;148:1-3. [PubMed]
17. Freedman AN, Seminara D, Gail MH, Hartge P, Colditz GA, Ballard-Barbash R, et al. Cancer risk prediction models: a workshop on development, evaluation, and application. J Natl Cancer Inst 2005;97:715-23. [PubMed]
18. Janssens ACJW, Ioannidis JPA, Bedrosian S, Boffetta P, Dolan SM, Dowling N, et al. Strengthening the reporting of genetic risk prediction studies (GRIPS): explanation and elaboration. PLoS Med 2011; doi:10.1371/journal.pmed.1000420.
19. Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663-94. [PubMed]
20. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med 2003;138:w1-12. [PubMed]
21. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JPA, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. BMJ 2009;339:b2700. [PMC free article] [PubMed]
22. Vandenbroucke JP, von Elm E, Altman DG, Gotzsche PC, Mulrow CD, Pocock SJ, et al. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. PLoS Med 2007;4:e297. [PMC free article] [PubMed]
23. Vandenbroucke JP. STREGA, STROBE, STARD, SQUIRE, MOOSE, PRISMA, GNOSIS, TREND, ORION, COREQ, QUOROM, REMARK... and CONSORT: for whom does the guideline toll? J Clin Epidemiol 2009;62:594-6. [PubMed]
24. Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation 2007;115:928-35. [PubMed]
25. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008;27:157-72. [PubMed]
26. Janssens ACJW, Khoury MJ. Assessment of improved prediction beyond traditional risk factors: when does a difference make a difference? Circ: Cardiovasc Genet 2010;3:3-5. [PubMed]
Articles from BMJ Open Access are provided here courtesy of
BMJ Group