Search Ebook here:


Oxford Textbook of Clinical Nephrology, 4th Edition



 PDF

Author: Neil Turner and David Goldsmith

Publisher: Oxford University Press

Genres:

Publish Date: December 22, 2015

ISBN-10: 199592543

Pages: 3040

File Type: PDF

Language: English

read download

Book Preface

Basic epidemiology principles in nephrology

Introduction

Epidemiology is the study of the distribution, determinants, and frequency of disease in populations or settings (Rothman, 1981, 2002). Therefore, epidemiological studies assess the extent of disease, risk/causal factors, natural history, prognosis, prevention/ treatment strategies, and the potential for new policies to prevent disease or improve outcomes (Rothman, 2002). Epidemiological research helps to inform evidence-based medicine by identifying risk factors for disease and to determine optimal treatment approaches; it is the cornerstone of public health research and of preventive medicine. The identification of unbiased causal relationships between exposures (risk factors or interventions) such as hypertension or the use of antihypertensive medication and outcomes like morbidity and mortality is therefore an important aspect of epidemiology. This section will discuss some epidemiological concepts, methods, and their application to clinical research in nephrology.

Research questions

Defining an appropriate research question requires familiarity with knowledge gaps in the subject area to judge if a question is feasible, interesting, novel, ethical, and relevant (FINER criteria) (Hulley et al., 2007). Whereas the FINER criteria highlight general aspects of research questions, the development of a specific research question may follow the PICO format, in which P stands for the population (or problem) of interest, I for the intervention (or any other exposure), C for the comparison group, and O for the outcome of interest. Some suggest a PICOT approach, adding a T for the follow-up time to assess outcome (Haynes et al., 2006). An example of a question framed according to PICOT is ‘In dialysis patients (P), what is the effect of statins (I) compared to placebo (C) on cardiovascular mortality (O) after 4 years of follow-up (T)?’ A research question has implications for the choice of the study design—which will in turn determine the analytical methods.

Study designs

Fig. 1.1 shows an algorithm for the classification of study designs in clinical research (Grimes and Schulz, 2002). Study design is an important aspect of study quality. Studies can be classified into experimental and observational ones depending on whether or not exposures like therapy were assigned by the investigators. Random allocation of exposures is important to prevent selection bias by the clinician (also known as ‘confounding by indication’) (Stel et al., 2007) occurring, when clinicians provide a specific therapy because of preconceived ideas about which therapy is best. Therefore, when it comes to studies on the effects of therapy or other interventions, randomized controlled trials (RCTs) are the gold standard. This was exemplified by RCTs showing a lack of effect or even harmful effects of using high target haemoglobin level in chronic kidney disease (CKD) patients as opposed to using a lower target (Drueke et al., 2006; Singh et al., 2006), whereas the majority of observational studies had suggested that higher haemoglobin levels were associated with favourable outcomes.

On the other hand, observational studies may answer questions on aetiology, diagnosis, prognosis, and adverse effects. In addition, they may provide answers on the effects of therapy where RCTs are not possible, inappropriate, inadequate, or unnecessary (Black, 1996). The effect of transplantation as compared to dialysis cannot be determined through an RCT, as allocation of renal grafts depends on other factors like HLA-matching. Where there is no comparison (control) group (as in case reports or case series), observational studies are called descriptive and where there is a comparison group they are referred to as analytical. Finally, the temporal direction of analytical observational studies determines the type of study. Cohort studies like the Dialysis Outcomes and Practice Patterns Study (DOPPS) determine the exposure of subjects to risk factors at the start of inclusion and then look forward in time to observe the occurrence of outcomes. They may provide a wealth of data which enable the investigator to study not only multiple outcomes but—unlike RCTs—also multiple exposures. In contrast, case–control studies compare cases (those with the disease or other outcome of interest) with controls (those without the outcome of interest) and then look back in time for exposures that might have caused the outcome. In nephrology, case–control studies are uncommon. Nevertheless, this study type is very efficient for studying potential risk factors for rare outcomes that may take a long time to develop, for example, CKD. By going back in time and looking for particular exposures like analgesics one may find associations between outcomes and these exposures. In such a case, prospective cohort studies are less efficient as one will need a very high number of subjects and a very long time to acquire an equal number of cases. Finally, cross-sectional studies examine the presence of an exposure and that of the outcome at the same moment in time. In most cases this simultaneity makes it difficult to determine which is the cause and which is the consequence, in other words, this design may induce a chicken-and-egg problem. Table 1.1 describes the strengths and weaknesses of different study designs (Jager et al., 2007).


Download Ebook Read Now File Type Upload Date
Download here Read Now PDF May 30, 2020

How to Read and Open File Type for PC ?