|
 |
| RESEARCH METHDOLOGY |
|
| Year : 2010 | Volume
: 1
| Issue : 1 | Page : 21-24 |
|
On research in clinical practice
Arun Nanivadekar
Formerly Medical & Research Director, Pfizer India.Currently, independent medical consultant at Mumbai and Pune
| Date of Web Publication | 20-Oct-2010 |
Correspondence Address:
Arun Nanivadekar
Formerly Medical & Research Director, Pfizer India.Currently, independent medical consultant at Mumbai and Pune
 Source of Support: None, Conflict of Interest: None  | Check |
PMID: 21829777 
 Abstract | | |
Clinical research implies advancing current knowledge about health care by continually developing and testing new ideas about diseases, products, procedures, and strategies. Although this trait is inherent in human nature, it needs to be encouraged, nurtured, groomed, and channelized by creating a suitable atmosphere for it, providing the necessary resources, inculcating the necessary conceptual and manual skills, and rewarding the efforts and achievements suitably. Language, logic, statistics, and psychology play an important role in acquiring and developing research capability. To be socially relevant and economically viable, clinical research will need to partner with patients and their doctors in identifying what their goals of health care are, what they value, and what they are willing to "buy" in terms of goods and services. Besides, clinical research will need to bring on one platform the sponsors, the researchers, the patients, the payers, and the regulators to ensure that they do not work at cross purposes, that the cost of developing health care measures is scaled down through innovative approaches such as large simple trials, sequential trials, early marketing conditional on post-marketing surveillance, and so on. All these will be possible if day-to-day practice is slowly and systemically transformed into the largest laboratory of clinical research, which it ought to be, by forming networks of research-oriented practices, and popularizing the use of data collection and analysis tools such as Epi Info which are in the public domain. Keywords: Clinical Research; Clinical Practice; Clinical Research Skills; Clinical Research Tools; Clinical Research Strategies
How to cite this article:
Nanivadekar A. On research in clinical practice. Perspect Clin Res 2010;1:21-4 |
Every medical practitioner is aware, I suppose, of the incompleteness and inadequateness of current medical knowledge. As a corollary, I believe he is also aware that there is a need to add to the existing knowledge, both by expanding it and by advancing it. This process of enhancing the existing knowledge is what research is expected to do. The process of research has two aspects: getting an idea is one, and verifying it is the other.
Getting an idea is the creative aspect of research. There is no sure way of how to do it. The person who is in the most advantageous position to get an idea is one who is constantly observing patients, diagnosing them, and treating them. During this work his sense organs constantly feed his mind with observations in which he might notice a trend, a pattern, a sequence, an association, or something not noticed or reported earlier. This process is intuitive. The idea appears to pop into the mind suddenly, but very likely it is the result of a subconscious process of reasoning that is going on in the depths of the mind for some time. Once formed, the idea may be communicated to the peers during a talk, or in writing as a case report, a case series, or a descriptive summary report.
Once the idea is outlined, it needs to be verified as to its truth or falsehood. This is the working aspect of research, requiring the making of a plan of how to test the idea; what information to collect, from whom, and how; and how to examine it to decide whether it supports the idea, contradicts it, or requires the idea to be modified. If we agree that this is how medical science has progressed in the past, and is likely to progress in the present and the future, then we need to ask who are in the most advantageous position to do it, and how they can be motivated and enabled to do it.
As I said earlier, constant exposure to events can spark off a chain of reasoning leading to an idea. Therefore, doctors who treat patients daily, i.e., practitioners, are in the most advantageous position to get ideas. History of medicine provides many examples of path-breaking ideas occurring to practicing doctors. How did Jenner think of inducing cowpox in people to protect them against smallpox? This is because he had heard from milkmaids that anyone who had cowpox before hardly ever got smallpox. His mind put two and two together. The facts that both infections are caused by viruses, and that they share antigenic similarity, became known much later.
Withering, who discovered the ability of digitalis to relieve heart failure; Lind, who discovered the usefulness of lemon juice for preventing scurvy; Hench, who discovered the efficacy of cortisol in relieving rheumatoid arthritis; all were practicing doctors working in the greatest laboratory of clinical research: the day-to-day world of sick people, at home, in consultation room, and in hospital.
Closer at home I recall an interesting incident. I was then following up trials of prazosin (an alpha-blocker) in hypertension, which had to be temporarily suspended because a few patients in the UK had fainted after taking the first dose. One of our investigators in Nagpur, a nephrologist, responded to the trial suspension by asking me what could be the likely reason for fainting. I replied that the suspected mechanism was venous dilation and pooling of blood in the visceral area. "I see.", he said, "If that is indeed so, why can we not use the drug to accelerate peritoneal dialysis in our patients who come from nearby towns and have to stay overnight? They are recumbent anyway, and we can monitor the CVP continuously. If we can reduce their dialysis time, we can send them home the same day." So he assigned this subject to a postgraduate as a thesis topic and, after six months or so, showed me the results of treating 20 patients in the conventional manner and 20 after giving them a dose of prazosin one hour before the procedure. The average time for dialysis was indeed halved, and the difference was real. Here I saw a nephrologist functioning as a classical clinical pharmacologist. Of course hemodialysis became commonplace soon after this event, the need of peritoneal dialysis became rare, and prazosin was not used or studied further for this purpose. Nonetheless, the originality of the idea was impressive.
I think the foregoing stories are enough to make the point that practicing doctors have the opportunity and ability to get new ideas that can change the current practice of medicine for the better. Let us now see how they can go about verifying the falsehood or truth of the ideas, and seek their acceptance by the peers. To do this, they need to be familiar with the scientific method and be trained in applying it to test different ideas. As Sir Sheldon Dudley has illustrated, [1] it has to do with language, logic, psychology, and statistics. This is one aspect that is neglected in the education and training of doctors, both during undergraduate and postgraduate years. And this is a gap which well trained and experienced pharmaceutical physicians can fill while working with practising doctors, either in the roles of postgraduate students, study coordinators, and investigators, or in the roles of research-oriented consultants and family physicians.
Any language is a system of symbols and sounds, and meanings associated with them. When we are thinking, we are quietly talking to ourselves. Words are the material with which we form and shape our ideas. So unless we acquire a reasonable mastery over language, and I do not mean only English, how can we think and communicate effectively? The ABC of communication - accuracy, brevity, clarity - are as much needed for thinking as for speaking and writing. Consider the writing part of doing research to test an idea. Be it the statement of an idea, a case report, a case series, or a summary report. Try to say or write what you mean, then ask the audience or the reader what he understood, and see how far the two are congruent. The results will surprise you.
Logic is concerned with thinking flawlessly, progressing from one statement to the next, until you are able to build a general concept from a particular case or cases, or draw an inference from statements presently accepted as facts. Both these processes are critical in research, i.e., in developing and testing ideas. Concepts such as association, causation, dependence, generalization, deduction, probability, plausibility, possibility are crucial for research.
Psychology also plays a role in research because human mind is its most important instrument. Knowing how it works, or can go astray, helps avoid pitfalls. Bias, prejudice, vested interests, envy, rivalry, and such attributes can influence thinking and the validity of research done.
Finally, statistics or the habit of quantifying what we observe, analyze, and infer. Numbers score over words because they do not have shades of meaning. Expressing facts and inferences in numbers is at the heart of testing ideas. The best way to learn this skill is by doing, which requires a self-learning tool and, if possible, a facilitator.
The tool I recommended for doctors keen to do research is a computer program called Epi Info [2] developed by the Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA. It can be used to design forms for recording data, organizing them, analyzing them, and presenting the results in tables or graphs. As regard a facilitator, I believe many pharmaceutical physicians will be happy to play this role as a part of their portfolio. Before I illustrate this with a few examples, I would like to state some beliefs which I have acquired during my industrial and academic career.
First, the questions which are important to patients, their doctors, and health care payers are fairly down-to-earth and concerned with benefits, risks, and costs of health care tools: drugs, devices, diagnostics, procedures, and strategies.
Second, the answer to such questions must be applicable to the broad category of patients seen in day-to-day practice, and not to a highly selective group of patients.
Third, we need large but simple databases to answer the question posed, which is possible if we collect just enough data on each patient. At present we seem to collect far more data on each patient than is really necessary, at least in clinical trials. To quote Peto et al., [3] "Most trials would be of much greater scientific value if they collected ten times less data, both at entry and during follow-up, and were therefore much larger."
Fourth, despite the availability of objective measures of illness and well-being, patients and doctors often judge the value of health care tools by what is subjectively important and relevant to them. The outcome of a health care measure may have many facets, and patients may assign different weights to them because of their job, family life, living conditions, social setting, finances, etc. [3] Rigid trials may drown this diversity into a uniform standard, but large, inpractice databases can integrate the same diversity into meaningful results. It is just like looking at a beam of light either through a prism or a plain glass: one will split the beam into different colors; the other will show it as it appears to all.
Let me take an example. You observe that the current epidemic of swine flu is leading patients to take an Ayurvedic or herbal medicine that is claimed to increase resistance to infection. If we have a database showing how many of those taking the medicine caught the infection, and how many of those who did not take the medicine caught it, we could at least have an idea about the likely benefit of taking it. If a group of practitioners could record the essential data - ID, initials, sex, age, locality, whether medication used (yes/no), and whether confirmed swine flu developed (yes/no), and any other medicine taken - in all patients they see during a defined period, we could have a meaningful answer. Suppose 20 practitioners recorded these data in 100 consecutive patients each, and found that 8 of 1100 who used the medicine got the illness (incidence 7 per 1000) whereas 19 of 900 who did not use the medicine got it (incidence 21 per 1000). Epi Info will help them record, organize, and analyze these data and conclude that the relative risk in the medication group was about a third (0.34) of that in the no-medication group, that the real relative risk could be in the range 0.15 to 0.78 (less than 1:1 or equal), that they can have 95% confidence in such a conclusion, and that the probability of these results occurring by chance is less than 1 in 100 (P < 0.01). Even if we do not know how the medicine works, at least taking it seems beneficial. If similar results are reported from various areas of the country, the validity of the conclusion will increase and perhaps prompt basic research into how the medicine might be working. Of course I have simplified the example. Even with such results, while interpreting them, one will have look for and rule out other possible causes for the difference (confounding factors). The article by Peipert and Phipps [4] provides a lucid and concise description of what a reliable observational study should be. One way to get familiar with this type of research is to work on the three interactive tutorials provided in the Help menu of the Epi Info program (shown below).[Additional file 2]
Another example that comes to my mind is that of a national survey of intestinal nematode infection which we organized in 1984 with the help of practicing pathologists. [5] We were then about to market the drug pyrantel, but my marketing colleagues were somewhat diffident because the drug had little efficacy against whipworm. From my discussions with practicing colleagues, I had developed the idea that if the prevalence of whipworm was very low and limited to a few pockets in the country, this fact itself would help dispel my colleagues' diffidence. Enlisting the help of our field staff, we listed the pathological laboratories in their territories (total 45) into three categories - public free hospital laboratories (PFHL), public paying hospital laboratories (PPHL), and private pathology laboratories (PPL) - on the reasoning that these serviced lower, higher, and middle socioeconomic strata, respectively. Using random numbers, we selected one laboratory of each category from each territory. We requested them to gives us anonymized copies of the first 100 consecutive stool examination reports beginning the first day of the next month. We only collected data on the presence or absence of roundworm, hookworm, and whipworm eggs. We thus had 45 x 3 x 100 = 13,500 stool reports in our database. Their analysis showed that the prevalence was 16.3% for roundworm, 14.7% for hookworm, and 3.7% for whipworm. Besides, whipworm was limited to three pockets: Mumbai, Kolkata, and Kerala state. These figures helped the marketing colleagues overcome their diffidence about competing against mebendazole, and gain confidence about the singledose convenience of pyrantel, especially for mass deworming, against the six-dose, three-day regimen of mebendazole.[Additional file 1]
One more example. The prevalence of sickle cell disease (SCD) is high in the backward areas of central India. There is no cure for it except bone marrow or stem cell transplant, which is not practicable for many patients. Hydroxyurea (HU) is a drug which promotes the formation of fetal hemoglobin, reduces the polymerization of hemoglobin S, and helps reduce the incidence and intensity of ischemic crises, and, if used long-term, can improve the quality of life and delay or prevent organ damage. However, its use is not adopted as widely as it should be. This may be due to ignorance, or unspoken fears about the possible side effects of HU which is otherwise known as an anticancer drug. Some trials [6] have shown that if used with simple hematological monitoring, the drug is well tolerated and offers considerable relief to SCD patients. Doctors practicing in the high prevalence area can form a network, use Epi Info to keep standardized records of their SCD patients, follow them up for years, and analyze the data for documenting the outcome among those who take and those who do not take HU. Such an observational, long-term study could help not only to quantify the benefits and risks of HU therapy, but also to create greater awareness of it among doctors and patients, and lead to its wider usage.
I shall rest my case for clinical research in practice with this eloquent quote: [7]
"The purpose of research is to create the knowledge essential for action to improve health. Without this knowledge, action is impossible because it has no logical or empirical* basis. Indeed, ongoing action for health, if it does not contain an imbedded program of research, frequently becomes irrelevant, misleading or unnecessarily costly. It is for this reason that the original conception of the National Tuberculosis Program (a forerunner of many other public health programs) included the necessity of an imbedded program of research as an essential component.
"Research is an activity of perpetual questioning. While public health practice is based on consensus, standardization and systematic practice, research requires a skeptical mind, prepared to continuously evaluate and question. This questioning and evaluating, when put into a systematic framework, creates the new knowledge that is required to create and continually modify actions for health. This is what research is and why it is important."
* The word empirical means "based on experience"; it does not mean irrational or arbitrary as some may believe.
| References | |  |
| 1. | Dudley, Sheldon. The Four Pillars of Wisdom. London: Watts & Co., 1947.  |
| 2. | www.cdc.gov/epiinfo |
| 3. | Peto R, Collins R, Gray R. Large scale randomized evidence: large, simple trials and overviews of trials. In, Warren KS, Mosteller F (ed). Doing More Good Than Harm - The Evaluation of Health Care Interventions. Annals of the New York Academy of Sciences 1993; 703: 314-340. |
| 4. | Peipert JF, Phipps MG. Observational studies. Clinical Obstetrics and Gynecology 1998; 41: 235-44. |
| 5. | Gadgil SD, Kulkarni SS, Apte VV, Nanivadekar AS. Intestinal nematode infections in India: a cross-sectional survey. Journal of Postgraduate Medicine 1984; 30: 137-43. |
| 6. | Italia K, Jain D, Gattani S, et al. Hydroxyurea in sickle cell disease - a study of clinico-pharmacological efficacy in the Indian haplotype. Blood Cells and Molecular Disease 2009; 42: 25-31. |
| 7. | Enarson DA, Kennedy SM, Miller DL, Bakke P. Research Methods for Promotion of Lung Health - A guide to protocol development for low-income countries. Paris: IUATLD, 2001; 11. |
|
