Академический Документы
Профессиональный Документы
Культура Документы
Conclusion
Clyde Williams is professor of sports science at the Fitness is a complex physiological characteristic that is difficult to
University of Loughborough.
The ABC of Sports Medicine has been edited by Greg describe comprehensively. Nevertheless, we can assess the central elements
McLatchie, visitingprofessor ofsports medicine and surgical of fitness in reliable and reproducible ways. Through assessing athletes
sciences at the University of Sunderland, consultant surgeon fitness we can extend health care by advising them on their ability to cope
at Hartlepool General Hospital, and director of the National with the exercise demands of their chosen sport.
Sports Medicine Institute, London.
The photographs of the javelin thrower (Tessa Sanderson) and of the woman gymnast-
(Jackie Brady) were taken by Supersport Photographs.
The "four principles plus scope" approach provides in committing ourselves to four prima facie moral
a simple, accessible, and culturally neutral approach principles plus a reflective concern about their scope of
to thinking about ethical issues in health care. The application. Moreover, these four principles, plus
approach, developed in the United States, is based attention to their scope of application, encompass most
on four common, basic prima facie moral commit- of the moral issues that arise in health care.
ments-respect for autonomy, beneficence, non- The four prima facie principles are respect for
maleficence, and justice-plus concern for their autonomy, beneficence, non-maleficence, and justice.
scope of application. It offers a common, basic "Prima facie," a term introduced by the English
moral analytical framework and a common, basic philosopher W D Ross, means that the principle is
moral language. Although they do not provide binding unless it conflicts with another moral principle
ordered rules, these principles can help doctors and -if it does we have to choose between them. The four
other health care workers to make decisions when principles approach does not provide a method for
reflecting on moral issues that arise at work. choosing, which is a source of dissatisfaction to people
who suppose that ethics merely comprises a set of
Nine years ago the BMJ allowed me to introduce to its ordered rules and that once the relevant information is
readers' an approach to medical ethics developed by fed into an algorithm or computer out will pop the
the Americans Beauchamp and Childress,' which is answer. What the principles plus scope approach
based on four prima facie moral principles and attention can provide, however, is a common set of moral
to these principles' scope of application. Since then I commitments, a common moral language, and a
have often been asked for a summary of this approach common set of moral issues. We should consider these
by doctors and other health care workers who find it in each case before coming to our own answer using our
Imperial College of helpful for organising their thoughts about medical preferred moral theory or other approach to choose
Science, Technology and ethics. This paper, based on the preface of a large between these principles when they conflict.
Medicine,
London SW7 INA multiauthor textbook on medical ethics,3 offers a brief
Raanan Gillon, visiting account of this "four principles plus scope" approach.
professor ofmedical ethics The four principles plus scope approach claims that Respect for autonomy
whatever our personal philosophy, politics, religion, Autonomy-literally, self rule, but probably better
BMY 1994;309:184-8 moral theory, or life stance, we will find no difficulty described as deliberated self rule-is a special attribute
Statistics Notes
Diagnostic tests 3: receiver operating characteristic plots
Douglas G Altman, J Martin Bland
This is the seventh in a series of We have previously considered diagnosis based on "curve" that coincided with the left and top sides of the
occasional notes on medical tests that give a yes or no answer.'2 Many diagnostic plot. A test that is completely useless would give a
statistics. tests, however, are quantitative, notably in clinical straight line from the bottom left corner to the top right
chemistry. The same statistical approach can be used corner. In practice there is virtually always some
only if we can select a cut off point to distinguish overlap of the values in the two groups, so the curve
"normal" from "abnormal," which is not a trivial will lie somewhere between these extremes.
problem. Firstly, we can investigate to what extent the A global assessment of the performance of the test
test results differ among people who do or do not have (sometimes called diagnostic accuracy4) is given by the
the diagnosis of interest. The receiver operating area under the receiver operating characteristic curve.
characteristic (ROC) plot is one way to do this. These This area is equal to the probability that a random
plots were developed in the 1950s for evaluating radar person with the disease has a higher value of the
signal detection. Only recently have they become measurement than a random person without the
commonly used in medicine. disease. (This probability is a half for an uninformative
We assume that high values are more likely among test-equivalent to tossing a coin.)
those dubbed "abnormal." Figure 1 shows the values No test will be clinically useful if it cannot
of an index of mixed epidermal cell lymphocyte discriminate,4 so a global assessment of discriminatory
reactions in bone marrow transplant recipients who did power is an important step. Having determined that a
or did not develop graft versus host disease.3 The test does provide good discrimination the choice can be
usefulness of the test for predicting graft versus host made of the best cut off point for clinical use. This
disease will clearly relate to the degree of non- overlap requires the choice of a particular point, and is thus a
between the two distributions. local assessment. The simple approach of minimising
A receiver operating characteristic plot is obtained "errors" (equivalent to maximising the sum of the
by calculating the sensitivity and specificity of every sensitivity and specificity) is not necessarily best.
observed data value and plotting sensitivity against Consideration needs to be given to the costs (not just
1-specificity, as in Figure 2. A test that perfectly financial) of false negative and false positive diagnoses
discriminates between the two groups would yield a and to the prevalence of the disease in the subjects
being tested.4 For example, when screening the general
10- 0 // /
population for cancer the cut off point would be chosen
to ensure that most cases were detected (high
8 0.8 S ,," sensitivity) at the cost of many false positives (low
8- i,0.6 ( specificity), who could then be eliminated by a further
test.
g0.4 A receiver operating characteristic plot is particularly
Medical Statistics useful when comparing two or more measures. A test
6- 0.2 with a curve that lies wholly above the curve of another
Laboratory, Imperial xw o000
0
Cancer Research Fund, 0 o will be clearly better. Methods for comparing the areas
London WC2A 3PX _ o 0 0.2 0.4 0.6 0.81 under two curves for both paired and unpaired data are
Douglas G Altman, head 4 - °00 - specificity reviewed by Zweig and Campbell,4 who give a full
assessment of this method.
Department ofPublic 00
00 000 FIGI1 (left) -Distribution of
Health Sciences, St 2 - 0 0 values ofan index ofmixed 1 Altman DG, Bland M. Diagnostic tests 1: sensitivity and specificity. BMJ
George's Hospital Medical 0 0° epidermal cell lymphocyte 1994;308:1552.
School, London nnn 0 reactions in patients who did or 2 Altman DG, Bland M. Diagnostic tests 2: predictive values. BMJ 1994;309:
did not develop graft versus host 102.
SW17 IRE 000000 3 Bagot M, Mary J-Y, Heslan M, et al. The mixed epidermal cell lymphocyte-
J Martin Bland, reader in 0 disease3 reaction is the most predictive factor of acute graft- versus-host disease in
medical statistics bone marrow graft recipients. BrJ Haematoll988;70:403-9.
No Yes FIG 2 (above)-Receiver 4 Zweig MH, Campbell G. Receiver-operating characteristic (ROC) plots: a
Graft v host disease operating characteristic curve for fundamental evaluation tool in clinical medicine. Clin Chem 1993;39:
BMJ 1994309:188 the data shown infig 1 561-77.