Bhandari / Jönsson Clinical Research for Surgeons

1. Auflage 2008
ISBN: 978-3-13-257926-2
Verlag: Thieme
Format: EPUB
Kopierschutz: 6 - ePub Watermark

E-Book, Englisch, 328 Seiten, ePub

Reihe: Princ. Pract. Clin. Res

is a practical guide for understanding, planning, conducting, and evaluating surgical research. It covers the principles of evidence-based surgery and applies these principles to the design of suitable research studies. The reader will come to fully understand important concepts such as case-control study, prospective cohort study, randomized trial, and reliability study. The book provides valuable discussions of the critical appraisal of published clinical studies, allowing the reader to learn how to evaluate the quality of such studies with respect to measuring outcomes and to make effective use of all types of evidence in patient care.

Highlights:

• Insights from experienced surgeons and veteran researchers
• Easy-to-reference text boxes with Key Concepts, Jargon Simplified, and Examples from the Literature
• Coverage of both open and minimally-invasive surgical procedures
• 50 illustrations demonstrating key points

This book is a valuable reference for clinicians and residents in a range of disciplines, including general surgery, orthopedic surgery, plastic and reconstructive surgery, urology, neurosurgery, otolaryngology-head and neck surgery, interventional radiology, cardiac surgery.

Zielgruppe

Ärzte

Autoren/Hrsg.

Bhandari, Mohit

Jönsson, Anders

Fachgebiete

• Medizin | Veterinärmedizin Medizin | Public Health | Pharmazie | Zahnmedizin Chirurgie Plastische, Rekonstruktive & Kosmetische Chirurgie
• Medizin | Veterinärmedizin Medizin | Public Health | Pharmazie | Zahnmedizin Chirurgie Neurochirurgie
• Medizin | Veterinärmedizin Medizin | Public Health | Pharmazie | Zahnmedizin Chirurgie Orthopädie- und Unfallchirurgie

Inhaltsverzeichnis

Part I Why We Need Clinical Research
1 Historical Perspective of Clinical Research
2 Evidence-Based Surgery Defined
3 Myths and Misconceptions about Evidence-Based Medicine
4 Becoming and Evidence-Based Surgeon
Part II Principles of Clinical Research
5 Various Research Design Classifications
6 Hierarchy of Research Studies: From Case Series to Meta-Analyses
7 Randomized and Nonrandomized Studies
Part IIA Understanding Research Study Design
8 The Clinical Case Series
9 The Case&ndsah;Control Study
10 The Prospective Cohort Study
11 The Randomized Trial
12 Meta-Analyses
13 Economic Analysis
14 The Diagnostic Study
15 The Reliability Study
Part IIB Understanding Outcomes Measurement
16 The Classification of Outcomes
17 What Makes an Outcome Measure Useful
18 Common Generic Outcome Scales for Surgeons
19 Common Disease-Specific Outcome Scales for Surgeons
Part IIC Understanding Treatment Effects
20 Common Ways to Present Treatment Effects
21 The Confidence Interval Defined
22 The P Value Defined
23 Errors in Hypothesis Testing
24 Clinical versus Statistical Significance
Part III Practice of Clinical Research
Part IIIA Planning a Research Study
25 The Requirements of a Clinical Research Proposal
26 The Identification of a Good Research Question
27 How to Conduct a Comprehensive Literature Search
28 Guide to Planning a Randomized Trial
29 Guide to Planning a Nonrandomized Study
30 Study Sample Size
31 How to Budget for a Research Study
32 Research Ethics, Review Boards, and Consent Forms
33 Regulatory Issues in the Evaluation of a New Device or Drug
34 Strategies in Research Funding
Part IIIB Conducting a Research Study
35 The Roles of the Research Team
36 The Role of a Central Methods Center
37 The Role of a Data Monitoring Committee
38 The Need for Separate Adjudication of Outcomes
39 Data Management
40 Study Case Report Forms
41 The Study Manual of Operations
Part IIIC Practical Guide to Statistical Analysis
42 Review of Basic Statistical Principles
43 Statistical Means and Proportions
44 Regression Analysis
45 Analysis of Variance
46 Correlation Defined

Leseproben

1

Historical Perspectives of Clinical Research

“.”

—

Summary

A brief history of clinical research, from the first human experiments to today's patient-centered, randomized controlled clinical trials, is presented.

Introduction

Clinical research has evolved over the centuries from basic controlled experiments to the sophisticated randomized, double-blind controlled clinical trials overseen by national safety and data monitoring boards. Although historically the adage, “the patient comes first” can be questioned, in this first decade of the 21st century there can be no doubt. International and national laws and standards have been put in place to guarantee the health, safety, and welfare of the patient taking part in clinical research. A brief overview of the historical process is presented here.

Controlled Experiments

Slotki describes a nutritional experiment involving a control group in the Book of Daniel from the Old Testament.1 The passage from Daniel describes not only a control group, but also a concurrent control group, a fundamental element of clinical research that gained wider acceptance only in the latter half of the 20th century. Although it may not be possible to confirm the accuracy of the account from the Old Testament, it is clear that the ideas existed around 150 BC when this passage was written. There appear to be no other recorded examples of thinking in comparative terms about the outcomes of medical treatment in ancient or medieval times. Lilienfeld provides an example from a fourteenth century the letter from Petrarch to Boccaceto, also cited by Witkosky.2,3

Petrarch believed that if a hundred or a thousand men of the same age, temperament, and habits, and in the same surroundings, were attacked at the same time by the same disease, and if one half followed the prescriptions of doctors and the other half took no medicine but relied on Nature's instincts, those treated by doctors were more likely to survive.

Packard4 describes the experiment of the renowned French surgeon, Ambroise Pare, who during a battle in 1537, applied the then recognized standard treatment for gunshot wounds and poured boiled oil over them. After running out of oil, he used a mixture of egg yolks, oil of roses, and turpentine. Upon visiting his patients the next day, he found that those who had received the new medication had little pain, their wounds had neither swollen nor become inflamed, and the patients had slept through the night. The patients who had received boiled oil, however, had fever with much pain and swollen wounds. He immediately abandoned the boiled oil treatment.

An early example of a randomized trial comes from the Belgian chemist Van Helmont.5 Van Helmont suggested taking 200 or 500 poor people that have fevers, pleurisies, and other ailments out of hospitals, camps, and elsewhere and dividing them into halves by casting lots –one half to be treated by him and the other half by another.

In 1747, Lind evaluated 12 patients with scurvy and their responses to various interventions. This is perhaps the most famous historical example of a planned, controlled, comparative, clinical trial.6 He found that those who ate oranges and lemons, natural sources of vitamin C, showed the most improvement. Similar comparative studies were conducted through the 1800s on the effect of drugs or vaccines in the treatment for smallpox, diphtheria, and cholera. Of note is the work of Pierre-Charles-Alexander Louis, a 19th-century clinician and pathologist, who introduced the applicability of statistics to medical practice.7 He pioneered the idea of comparing the results of treatments on groups of patients with similar degrees of disease, that is, “like with like.”

In 1863 Austin Flint (1812–1886) conducted the first-ever trial that directly compared the efficacy of a dummy simulator (a placebo) with that of an active treatment; although Flint's examination did not compare the two against each other in the same trial. Even so, this was a significant departure from the (then) customary practice of contrasting the consequences of an active treatment with what Flint described as “the natural history of [an untreated] disease”.8

The literature of the 1900s primarily consisted of studies focused on prophylaxis and the treatment of infectious diseases. However, in 1947, the design of clinical trials entered a new, progressively more scientific phase spear-headed by the British Medical Research Council (MRC), with the first placebo-controlled randomized clinical trial, which evaluated the effect of streptomycin on tuberculosis. This was the first formal study to randomly assign patients to experimental and control groups.

Randomized Trials

Randomized trials date back to at least 1898 when Fibiger used systematic assignment to assign diphtheria patients to serum treatment or an untreated control group.9 Amberson and McMahon in 1931 also used group randomization in a trial of sanocrysin for the treatment of pulmonary tuberculosis.10 However, this type of alternating assignment was discredited shortly thereafter: patients’ knowledge of their treatment protocol introduced selection bias and called into question the treatment results.11

Obviously, new methods in how clinical investigators assigned patients to treatment groups were needed to offset bias and the subsequent invalidation of study data.12 In the 1947 MRC study mentioned above, Bradford Hill used random sampling numbers in assigning treatment to subjects in the study of streptomycin in pulmonary tuberculosis.13 Under the auspices of the MRC, he continued with further randomized trials: chemotherapy of pulmonary tuberculosis in young adults,14 antihistamine drugs in the prevention and treatment of the common cold,15 cortisone and aspirin in the treatment of early cases of rheumatoid arthritis,16,17 and long-term anticoagulant therapy in cerebrovascular disease.18 These early randomized clinical trials helped to transform clinical research into a rigorous science with formal methods that minimize investigator bias, design flaws, and subjective evaluation of treatment data.

The U.S. National Institutes of Health (NIH) started its first randomized trial in 1951, which was a study comparing adrenocorticotropic hormone (ACTH), cortisone, and aspirin in the treatment of rheumatic heart disease.19 This was followed in 1954 by a randomized trial of retrolental fibroplasias (now known as retinopathy of prematurity).20 Since World War II, the prospectively randomized controlled clinical trial has become the gold standard for evaluating new practices and therapeutic agents in medicine. In their 1976 and 1977 companion articles, Peto et al21,22 provide a thorough description of the design and analysis of modern clinical trials.

Blind Studies

An important factor in the success of a clinical trial is the avoidance of any bias in the comparison of the groups. “Blinding” the investigator, patient, and assessor prevents an outcome assessment bias. The randomization of patients avoids possible bias upon treatment allocation; nevertheless, bias can creep in while the study is in process. Both the patients and the doctor may be affected by their treatment response and knowledge of treatment given, respectively. For this reason, neither the patient nor the person evaluating the patient should know which treatment is given. In this case, the trial is called a study. If only the patient is unaware of the treatment given, the trial is called a study. In several fields, including surgery, it is often impossible for a study to be double-blinded; nevertheless, all trials should use the maximum degree of blindness that is possible.

Once again, it was Bradford Hill in the MRC landmark trial investigating the effect of streptomycin on tuberculosis who introduced the concept of blinding. In the study, two radiologists and a clinician each read the study patients’ radiographs independently and were unaware of whether the films were of “C” (control, bed rest alone) or “S” (streptomycin and bed rest) cases. Recognizing that if blinding and randomization were to be used effectively, Bradford Hill tried to ensure that judgments were made without any possible bias, without any overcompensation for any possible bias, and .23

Sequential Trials

In a sequential trial, each participant's results are analyzed after data become available. If the superiority of a treatment is established, if there are adverse treatment effects, or if it is determined that a treatment difference is unlikely, the study may be terminated. However, sequential trials are best used when the outcome of interest is quickly known. A sequential trial promotes patient safety and treatment benefits, as well as ensuring study cost-effectiveness. In the group sequential trial, the frequency of interim analyses is usually limited to between three and six.

Bross and Armitage acknowledged the need for interim analysis and the recognition that such analyses affect the probability of the type 1 error with the publication in the 1950s of papers on sequential clinical trials.24,25 The main...