About OMICS Group International is an amalgamation of
About OMICS Group International is an amalgamation of Open Access publications and worldwide international science conferences and events. Established in the year 2007 with the sole aim of making the information on Sciences and technology ‘Open Access’, OMICS Group publishes 400 online open access scholarly journals in all aspects of Science, Engineering, Management and Technology journals. OMICS Group has been instrumental in taking the knowledge on Science & technology to the doorsteps of ordinary men and women. Research Scholars, Students, Libraries, Educational Institutions, Research centers and the industry are main stakeholders that benefitted greatly from this knowledge dissemination. OMICS Group also organizes 300 International conferences annually across the globe, where knowledge transfer takes place through debates, round table discussions, poster presentations, workshops, symposia and exhibitions.
About OMICS Group Conferences OMICS Group International is a pioneer and leading science event organizer, which publishes around 400 open access journals and conducts over 300 Medical, Clinical, Engineering, Life Sciences, Pharma scientific conferences all over the globe annually with the support of more than 1000 scientific associations and 30, 000 editorial board members and 3. 5 million followers to its credit. OMICS Group has organized 500 conferences, workshops and national symposiums across the major cities including San Francisco, Las Vegas, San Antonio, Omaha, Orlando, Raleigh, Santa Clara, Chicago, Philadelphia, Baltimore, United Kingdom, Valencia, Dubai, Beijing, Hyderabad, Bengaluru and Mumbai.
. S. , C i h jyot a t of y n g e a h , tm Dr. B , Depar entistry der ealth D and a e e g R e l lic H tal Col e, Pub en D nger i j a u v a p Ba ital, D p dia. n I Hos , taka a n r Ka A Practical G uide To The S tatistical Consideration s Of Bio-medi cal Research.
Statistics is an essential and integral part of research methodology Methods of statistical analysis are absolutely indispensable and are powerful tools providing the techniques for drawing objective and scientific conclusions for the investigators. An amazing tool in communicating the necessary information about any research activity Very vital to the biomedical research as it constitutes the integral part of medical and dental curricula A pervasive force on which the entire spectrum of clinical decision making is dependent upon. Accurate and reliable statistics and its correct use are essential to a high quality study. Conclusions based upon poor research or its misinterpretation due to inappropriate statistics can be detrimental to patient care. 9/25/2020 The results of the published work influence further research by other researchers and so have a strong impact on the clinical decision making. Statistical considerations of biomedical research 1
Thereby, misuse or incorrect use of statistical methods may point the research in the wrong direction. The poor quality of statistics in published papers has been a cause of concern for many years, and is not confined to medical research. Furthermore, papers that cause the most trouble are usually those using only simple statistical methods rather than those with more complicated analysis. At times, it is impossible to assess the seriousness of many of the errors committed in research, as because an invalid analysis may give the same result as an appropriate one. Unwantedly, many published scientific articles have shortcomings in statistical design and analysis. Statistical errors are subtle, technical and difficult to detect, while the impact of statistical errors is not easy to measure. Consequently, the research conclusions based on misused statistics are slow to show their effects. “There are three kinds of lies: lies, damned lies and statistics” - Benjamin Disraeli. 9/25/2020 Statistical considerations of biomedical research 2
1. Ignorance towards the type of analysis at the planning stage. 2. Failure to determine the sample size scientifically. 3. Selective or self selection of samples. 4. Not paying adequate attention to the assumptions underlying the statistical tests. 5. Chasing the P value. 6. Multiple group comparisons without paying due importance to the inflation of Type I error. 7. Concealment of neutral and negative results. 9/25/2020 Statistical considerations of biomedical research 3
1. Ignorance towards the type of analysis at the planning stage of the study. Scale of measurement of data • Continuous scale (Interval or ratio) imparts high power and also smaller sample size. • Parametric tests – most powerful and robust. • Influences sample size calculation. Selection of statistical tests Generalizability One must decide upon the statistical approach to analyzing the data even before the data collection and sample size Statistical considerations of biodetermination. 9/25/2020 medical research 4
2. Failure to determine the sample size scientifically. Must account for the non-responders and drop outs. Sample size Undersized sample -Primary factor lacks power to detect researchers use to the possible control the power of a effect – inconclusive. study. Oversized sample – cumbersome and unethical, makes even a trivial effect to turn significant. Concept of saturation- guiding principle for qualitative studies. ‘Larger the sample size the better the study’ is not always true. 9/25/2020 Statistical considerations of biomedical research 5
3. Selective or self-selection of samples. May invalidate statistical tests. Selective or self -selection of samples. Leads to selection bias- hinders generalizability. Careless sampling is much more of a problem than an inappropriate analysis. Sample size estimation may not hold good. 9/25/2020 Statistical considerations of biomedical research 6
4. Not paying adequate attention to the assumptions underlying the statistical tests. Yes Normal distribution (Function of sample size) Parametric tests Log transformation No Nonparametric tests “Never replace parametric tests for non-parametric tests”. 9/25/2020 Statistical considerations of biomedical research 7
5. Chasing the ‘P’ value Statistical significance (P value) Measure of chance occurrence. 9/25/2020 Statistical considerations of biomedical research Clinical significance Measure of clinically important effect. 8
5. Chasing the ‘P’ value Larger value could be a result of smaller sample size. To be presented along with estimates of effect and confidence Intervals. P value Exact value is preferred instead of standard ones (P ≤ 0. 05, P≤ 0. 01 & P ≤ 0. 001). Smaller value could be a result of larger sample size. 9/25/2020 Statistical considerations of biomedical research 9
6. Multiple group comparisons without paying due importance to inflation of type I error. • Multiple hypotheses (especially post hoc) or subgroups are tested at the end of a trial or during interim testing while the trial is in progress. • Also known as fixing expedition or data dredging. • The more we look for a difference the more likely we are of finding one even by chance. Multiple hypotheses testing. 9/25/2020 Inflation of type I error. • For every 20 hypotheses tests that are performed one significant finding will emerge just by chance. • Can be compensated for by means of Bonferroni correction but that in turn will place additional demand on the sample size to achieve statistical significance at the corrected level of significance. Number of pairwise comparisons for a n - group study is given by n(n-1)/2 Statistical considerations of biomedical research • Constrain ourselves to a very specific primary hypothesis that can be tested statistically without having to correct for multiple testing. • Perform only a minimal number of preformed comparisons. • State a secondary hypothesis concerning the research question in advance and perform post-hoc analysis on the most statistically important findings. Practical solution. 10
Number of comparisons Probability of Type I error 9/25/2020 1 5% 2 10% 3 4 5 6 7 8 9 10 14% 19% 23% 27% 30% 34% 37% 40% Statistical considerations of biomedical research 11
7. Concealment of neutral and negative results. Manipulation of the study findings – unethical. Use of medical Rx that is ineffective, unpleasant, costly or even dangerous. Truth is not only violated by falsehood; Neutral and Negative results. Wastage of resources in replicating the same findings. Likely to face rejection by the editors or reviewers. Publication bias. Failing to report the findings -“File -drawer effect” Major threat to the validity of systematic reviews and meta-analyses. 9/25/2020 Statistical considerations of biomedical research it may be equally outraged by silence. Henri Frederic Amiel, 1821 -1881. 12
Statistical implications to bio-medical research. 9/25/2020 Statistical considerations of biomedical research 13
Recommendations for improving the quality of statistical aspects of bio-medical research. Re-orientation of journal editorial boards. Employ a statistician. Quality- a criteria for publication. 9/25/2020 Re-organization of educational institutions actively involved in bio-medical research. Institutional review boards(IRB) to have a statistically competent personnel. Mentors to bring in orientation towards applied statistics among… Students. Clinicians. Appoint a statistician. Statistical considerations of biomedical research 14
Conclusion………. No research is perfect in itself. Focus should be on curtailing the statistical inadequacies in any research to the best possible level within practical limits. Besides, Nothing can be a substitute for the knowledgeable interpretation of data. Hence using computing power should go hand in hand; not replace statistical reasoning. 9/25/2020 Statistical considerations of biomedical research 15
Dr. K. Sadashiva shetty, Principal, Bapuji Dental College and Hospital. Karnataka, India. Dr. L. Nagesh. , Head of the Department, Department of Public Health Dentistry, Institute of Dental Sciences, Bareilly, Uttar Pradesh, India.
- Slides: 22