Review articles
 

By Dr. Brijesh Sathian , Dr. Jayadevan Sreedharan
Corresponding Author Dr. Brijesh Sathian
Community Medicine, Manipal College of Medical Sciences, Department of Community Medicine, Manipal College of Medical Sciences - Nepal 155
Submitting Author Dr. Brijesh Sathian
Other Authors Dr. Jayadevan Sreedharan
Research Division, Gulf Medical University, Assistant Director & ProfessornResearch Division, Gulf Medical University - Nepal 155

BIOSTATISTICS

Medical Research, Testing of Hypothesis, p-value

Sathian B, Sreedharan J. Meaning of p-value in Medical Research. WebmedCentral BIOSTATISTICS 2012;3(5):WMC003338
doi: 10.9754/journal.wmc.2012.003338

This is an open-access article distributed under the terms of the Creative Commons Attribution License(CC-BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
No
Submitted on: 06 May 2012 05:57:50 PM GMT
Published on: 07 May 2012 04:21:33 PM GMT

Abstract


Any researcher begins the research with null hypothesis and alternative hypothesis. Null will be for supporting the old fact and alternative will be for the new fact invented/ doubted by the researcher/ scientist. Next step is to select one of this scientifically by using the science of statistics. For that the researcher should calculate the likelihood or probability that the difference observed in the study, however big or small, could have arisen purely by chance. This probability is known as p-value and it is sufficiently small, you can conclude that you have obtained a statistically significant difference. Confidence intervals and p-values take as their starting point the results observed in a study. Crucially, we must check first that this is an unbiased study.

Introduction


“Medicine is a science of uncertainty and an art of probability”, mused William Osler.  Medical journals are a confluence of medicine, science and journalism—and are expected to have the values of all three. Medical journals differ fundamentally from scientific journals in that the former is read mainly by practising doctors and not by scientists. Medical journals will continue to be the main vehicle of scientific information for years to come, particularly where access to computer and internet facilities are relatively limited. Nowadays the output—and rewards— of research are based almost entirely on published papers in scientific journals. Scientists in low-income and middle income settings want an opportunity to analyze data for their populations according to their own concerns. They want to be in the frontlines of national and global communications about their country’s experiences. The basic assumption of inferential statistics is that we’re observing a sample of finite size drawn from a population that is effectively infinite. By making observations about the sample, we are trying to make generalizations about the population. The Neyman-Pearson theory of hypothesis testing addresses the problem of choosing between two statistical hypotheses, H0 and H1. The solution involves selecting, before the data are observed, a set of potential observations (the critical region), then choosing H1 if the actual observations fall in the critical region, and choosing H0 if they do not. More clearly in large sample test, we will check the calculated value of the test formula is greater than the tabular value from the Z- table. Value is 1.96 for 95% and 2.58 for 99% Confidence Interval. The other method is so called software adopted method or widely used method p- value. It is described by Cox and Hinkley. Several people described p-value in several manners but finally it is the supporting probability of Null Hypothesis [1-30].

A p-value is a measure of how much evidence we have against the null hypothesis. The null hypothesis H0, represents the hypothesis of no change or no effect, population mean and sample mean equal, population proportion or sample proportion equal. It is also known as producer’s hypothesis, if customer is doing research. The smaller the p-value, the more evidence we have against H0. It is also a measure of how likely we are to get a certain sample result or a result more extreme, assuming H0 is true.  Medical research involves making a hypothesis and then collecting data to test that hypothesis. The p-value measures consistency by calculating the probability of observing the results from your sample of data or a sample with results more extreme, assuming the null hypothesis is true. We will reject H0 when p-value is less than 0.05. Sometimes, though, researchers will use a stricter cut-off (e.g., 0.01) or a more liberal cut-off (e.g., 0.10). The general rule is that a small p-value is evidence against the null hypothesis while a large p-value means little or no evidence against the null hypothesis.

Suppose that a Viagra medicine company alleges that only 40% of all patients who take it have a side effect of headache. If, you prescribed Viagra for effective treatment for Erectile Dysfunction and believe that the adverse event rate is much higher in patients. In a sample of 100 patients, all hundred have a headache. The data supports your belief because it is inconsistent with the assumption of a 40% headache rate. It would be like tossing a coin 100 times and getting heads each time. The p-value, the probability of getting a sample result of 100 headache events in all 100 patients assuming that the headache event rate is 40%, is a measure of this inconsistency. The p-value, 0.00000001, is small enough that we would reject the hypothesis that the headache event rate was only 40%. Researcher see a large p- value, he will decide null hypothesis is wrong. But if the sample size is not adequate and the sample is not the proper representative part then automatically p value will be high even though null hypothesis clinically right. So, the researcher should also look for one of two things: First one is the power of the test which confirms that the sample size in that study was adequate for detecting a clinically relevant difference. Second one is confidence interval that lies entirely within the range of clinical indifference. Researcher should also be cautious about a small p-value, but for different reasons. In some situations, the sample size is so large than required then the researcher can prove the medically insignificant result statistically significant [1-30].

Consultant is reading a research paper for finding a better drug for a particular disease. In a good research paper author of the research paper should inform you what size difference is clinically relevant and what sized difference is trivial. But all the manuscripts are not good. So, the reader has to find out how much of a difference would be large enough. Then compare this to the confidence interval in the research paper. If both limits of the confidence interval are smaller than a clinically relevant difference, then consultant should not change the drug, no matter what the p-value tells.

Conclusion(s)


P-value is a very important concept for any medical researcher for understanding the research articles published and for making correct statistical and medical inferences in his own paper. Researcher should not interpret the p-value as the probability that the null hypothesis is true. Such an interpretation is problematic because a hypothesis is not a random event that can have a probability. 

References


1. Davies HT. Interpreting measures of treatment effect. Hosp Med 1998; 59: 499–501.
2. Guyatt  GH, Sackett DL, Cook DJ. Users’ guides  to the medical literature. II. How to use an article about therapy or prevention. A. Are the results  of the study  valid? JAMA 1993; 270: 2598–2601.
3. Sackett DL, Haynes  RB, Guyatt  GH, Tugwell P. Clinical epidemiology: A basic science for clinical medicine, 2nd  edn. Boston,  Massachusetts: Little, Brown and Company, 1991.
4. Sackett DL, Richardson WS, Rosenberg W, Haynes  RB. Evidence based medicine: how to practice and teach EBM. London: Churchill Livingstone, 1997.
5.  Crombie  IK. The  pocket  guide  to  critical  appraisal. London: BMJ Publishing, 1996.
6. Brennan P, Croft P. Interpreting the results  of observa- tional research: chance is not  such  a fine  thing. BMJ 1994; 309: 727–730.
7. Burls A. What is critical appraisal? London: Hayward Medical Communications, 2009.
8. Gardner MJ, Altman  DG. Confidence intervals rather than p values: estimation rather than hypothesis testing. BMJ 1986; 292: 746–750.
9. Last JM. A dictionary of epidemiology. Oxford: International Journal of Epidemiology, 1988.
10.  The  Heart  Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting- enzyme inhibitor, ramipril, on cardiovascular events  in high-risk patients. N Engl J Med 2000; 342: 145–153.
11. Altman  DG, Bland JM. Absence of evidence is not  evi- dence of absence. BMJ 1995; 311: 485.
12. Guyatt  GH, Sackett DL, Cook DJ. Users’ guides  to the medical literature. II. How to use an article about therapy or  prevention. B. What  were  the  results  and  will  they help  me  in  caring  for  my  patients? JAMA 1994;  271:59–63.
13. Mittal A ,  Sathian B, Chandrasekharan N , Lekhi A, Farooqui M S, Pandey N. Diagnostic Accuracy of Serological Markers in Viral Hepatitis and Non Alcoholic Fatty Liver Disease. A Comparative Study in Tertiary Care Hospital of Western Nepal. Nepal Journal of Epidemiology 2011;1(2): 60-3.
14. Mittal A, Sathian B, Kumar A, Chandrasekharan N, Dwedi S. The Clinical Implications of Thyroid Hormones and its Association with Lipid Profile: A Comparative Study from Western Nepal. Nepal Journal of Epidemiology 2010; 1(1): 11-6.
15. Sathian B. Methodological Rigors in Medical Journals from Developing Countries: An Appraisal of the Scenario in Asia. Nepal Journal of Epidemiology 2011; 1(5): 141-43.
16. Sathian B, Sreedharan J, Mittal A, Chandrasekharan N, Baboo NS, Abhilash ES, et.al. Case Control Studies in Medical Research. Nepal Journal of Epidemiology 2011;1(3): 77-8.
17. Sathian B. Reporting dichotomous data using Logistic Regression in Medical Research: The scenario in developing countries. Nepal Journal of Epidemiology 2011;1(4):111-113.
18.  Sathian B, Sreedharan J, Baboo NS, Sharan K, Abhilash E S, Rajesh E. Relevance of Sample Size Determination in Medical Research. Nepal Journal of Epidemiology 2010; 1(1): 4-10.
19. Roy B, Banerjee I, Sathian B, Mondal M, Saha CG. Blood Group Distribution and Its Relationship with Bleeding Time and Clotting Time: A Medical School Based Observational Study among Nepali, Indian and Srilankan Students. Nepal Journal of Epidemiology 2011;1(4):135-40.
20. Sreeramareddy CT, Ramakrishnareddy N, Harsha KumarHN, Sathian B, Arokiasamy JT. Prevalence, distribution andpredictors of tobacco smoking and chewing in Nepal: a secondary data analysis of Nepal Demographic and HealthSurvey-2006. Substance Abuse Treatment, Prevention, and Policy 2011;6:33.
21. Roy B, Banerjee  I, Sathian B, Mondal M, Kumar SS, Saha CG. Attitude of Basic Science Medical Students towards  Post Graduation in Medicine and Surgery: A Questionnaire based Cross-sectional Study from Western Region of Nepal. Nepal Journal of Epidemiology 2010; 1(4):126-34.
22. Banerjee I, Roy B, Sathian B, Banerjee I, Kumar SS, Saha A. Medications for Anxiety: A Drug utilization study in Psychiatry Inpatients from a Tertiary Care Centre of Western Nepal. Nepal Journal of Epidemiology 2010; 1(4):119-25.
23. Mittal A, Sathian B, Kumar A, Chandrasekharan N, Farooqui MS, Singh S, Yadav KS. Hyperuricemia as an Additional Risk Factor for Coronary Artery Disease: A Hospital Based Case Control Study in Western Region of Nepal. Nepal Journal of Epidemiology 2011;1(3):81-5.
24. Basha AS, Mathew E, Sreedharan J, Muttappallymyalil J, Sharbatti AS, Shaikh BR. Pattern of Blood Pressure Distribution among University Students in Ajman, United Arab Emirates. Nepal Journal of Epidemiology 2011;1(3):86-9.
25. Banerjee I, Jauhari AC, Bista D, Johorey AC, Roy B, Sathian B. Medical Students View about the Integrated MBBS Course: A Questionnaire Based Cross-sectional Survey from a Medical College of Kathmandu Valley. Nepal Journal of Epidemiology 2011;1(3): 95-100.
26. Mittal A, Sathian B, Poudel B, Farooqui MS, Chandrasekharan N, Yadav KS. The Significance of Hepatobiliary Enzymes for Differentiating Liver and Bone Diseases: A Case Control Study from Manipal Teaching Hospital of Pokhara Valley. Nepal Journal of Epidemiology 2011;1(5): 153-9.
27. Poudel B, Mittal A, Yadav BK, Sharma P, Jha B, Raut KB. Estimation and Comparison of Serum Levels of Sodium, Potassium, Calcium and Phosphorus in Different Stages of Chronic Kidney Disease. Nepal Journal of Epidemiology 2011;1 (5): 160-7.
28. Mittal A, Sathian B, Chandrasekharan N, Lekhi A, Rahib R,  Dwedi S. Hepatic Steatosis and Diabetes Mellitus: Risk Factors, Pathophysiology and with its Clinical Implications: A Hospital Based Case Control Study in Western Region of Nepal. Nepal Journal of Epidemiology 2011;1(2):51-56.
29. Banerjee I, Roy B, Banerjee I, Sathian B, Mondol M, Saha A. Depression and its Cure : A Drug Utilization Study from a Tertiary Care Centre of Western Nepal. Nepal Journal of Epidemiology 2011;1 (5):144-52.
30. Mittal A, Sathian B, Kumar A, Chandrasekharan N, Sunka A. Diabetes mellitus as a Potential Risk Factor for Renal Disease among Nepalese: A Hospital Based Case Control Study. Nepal Journal of Epidemiology 2010; 1(1): 22-5.

Source(s) of Funding


none

Competing Interests


none

Disclaimer


This article has been downloaded from WebmedCentral. With our unique author driven post publication peer review, contents posted on this web portal do not undergo any prepublication peer or editorial review. It is completely the responsibility of the authors to ensure not only scientific and ethical standards of the manuscript but also its grammatical accuracy. Authors must ensure that they obtain all the necessary permissions before submitting any information that requires obtaining a consent or approval from a third party. Authors should also ensure not to submit any information which they do not have the copyright of or of which they have transferred the copyrights to a third party.
Contents on WebmedCentral are purely for biomedical researchers and scientists. They are not meant to cater to the needs of an individual patient. The web portal or any content(s) therein is neither designed to support, nor replace, the relationship that exists between a patient/site visitor and his/her physician. Your use of the WebmedCentral site and its contents is entirely at your own risk. We do not take any responsibility for any harm that you may suffer or inflict on a third person by following the contents of this website.

Reviews
1 review posted so far

Meaning of p-value in Medical Research
Posted by Dr. William J Maloney on 10 Feb 2014 09:56:00 PM GMT Reviewed by Interested Peers

Comments
1 comment posted so far

p-value and medical research Posted by Dr. Brijesh Sathian on 07 May 2012 05:50:41 PM GMT

Please use this functionality to flag objectionable, inappropriate, inaccurate, and offensive content to WebmedCentral Team and the authors.

 

Author Comments
0 comments posted so far

 

What is article Popularity?

Article popularity is calculated by considering the scores: age of the article
Popularity = (P - 1) / (T + 2)^1.5
Where
P : points is the sum of individual scores, which includes article Views, Downloads, Reviews, Comments and their weightage

Scores   Weightage
Views Points X 1
Download Points X 2
Comment Points X 5
Review Points X 10
Points= sum(Views Points + Download Points + Comment Points + Review Points)
T : time since submission in hours.
P is subtracted by 1 to negate submitter's vote.
Age factor is (time since submission in hours plus two) to the power of 1.5.factor.

How Article Quality Works?

For each article Authors/Readers, Reviewers and WMC Editors can review/rate the articles. These ratings are used to determine Feedback Scores.

In most cases, article receive ratings in the range of 0 to 10. We calculate average of all the ratings and consider it as article quality.

Quality=Average(Authors/Readers Ratings + Reviewers Ratings + WMC Editor Ratings)