The Measurement of Treatment Effects

A disease may have many symptoms and a treatment may have many effects. A particular disease may affect individuals differently and so may a given treatment. We do not study all individuals to whom we may eventually apply the treatment but a rather unrepresentative subset who form the patients in our clinical trial. We should like to be able to use the experience gained with these patients to say something about future patients. Measuring to the best extent that we are able what a particular treatment does is not the same as measuring how well it does what we should like it to do. Furthermore, certain standard statistical tools require measurements of a particular form. All of these are reasons why the measurement of effects in a clinical trial is not necessarily a simple matter. They are also reasons why measurement is not just the concern of the physician but also the concern of the statistician. 

Some of the issues have already been touched on in Chapters 3 and 7 others will be dealt with when we come to consider multiplicity in Chapter 10, and also safety of treatments in Chapter 23. (In particular, some matters regarding survival analysis will be touched on in that chapter.) Other remarks and issues to do with measurement will be found in various chapters on more particular specialist matters for example. Chapter 21 which deals with pharmacokinetics and Chapter 24 which deals with pharmaco-economics. In this chapter we shall cover those important Issues which remain. 

Statistical Issues in Drug Development, 2nd Edition. Stephen Senn 2007 John Wiley Sons, Ltd. ISBN 978-0-470-01877-4 

If the effect of a given treatment for asthma is to increase forced expiratory volume in one second (FEV ) 12 hours after treatment by 300 ml above what it would have been without treatment, (whatever it would have been), then the treatment is additive on the FEVi scale. If it increases FEV by 15%, however, then while it increases a value which would have been 2000 ml by 300 ml, it only increases a value which would have been 1500 ml by 225 ml, and one which would have been 1000 ml is only increased by 150 ml, Thus, treatment is not additive on the FEV scale. However, if we take logarithms of the data then we can restore additivity, as is illustrated in Table 8.1. 

Another transformation which is commonly employed by statisticians is the logit transformation. Suppose we are interested in looking at the effect of a treatment on the probability of survival of patients over a given time period. The probability of survival will lie between 0 and 1 for any patient. If we use the probability scale to make our analysis we may come to some conclusion such as (say) the effect of treatment is to increase the probability of survival by 0.23. Suppose, however, that we now wish to apply the treatment to a type of patient whose probability of survival without treatment we believe to be 0.86. Applying our treatment estimate would lead to the nonsensical conclusion that her chances of survival were now 0.86 + 0.23 = 1.09 This can be avoided if, instead of using a scale like the probability scale, which is bounded by 0 and 1, we use a scale which, although related to it, is not so bounded. An example of such a scale is the logit scale and it is defined by 

---

In Table 8.5, we compare the response rates for the two primary endpoints - disease deterioration and mortality for the Flindle et al. study. What is interesting is that for the mortality endpoint ARR shows less deviation from the null than in the case of disease deterioration, while the converse holds for the RR. This is often regarded as a major defect of the RR as a measure of treatment effect, in that it does... 

Laupacis et al. introduced the number needed to treat (NNT) into the medical literature" as an easily imderstood and useful measure of treatment effect for clinical trials in which the main outcome variable is binary. It has been argued that the NNT is more easily... 

Each trial that is to be included in the meta-analysis will provide a measure of treatment effect (difference). For continuous data this could be the mean response on the active treatment minus the mean response in the placebo arm. Alternatively, for binary data the treatment effect could be captured by the difference in the cure rates, for example, or by the odds ratio. For survival data, the hazard ratio would often be the measure of treatment difference, but equally well it could be the difference in the two-year survival rates. 

Cook RJ, Sackett DL. The number needed to treat. A clinically useful measure of treatment effect. BMJ 1995 310 452-4... 

Acion L, Peterson JJ, Temple S, Arndt S (2006) Probabilistic index an intuitive non-parametric approach to measuring the size of treatment effects. Statistics in Medicine 25 591-602. 

Figure 1 shows the lack of treatment effects immediately following salinization as well. In situ gas exchange measurements were made daily in early afternoon. The plants at the higher salinity were increased from 10 to 50 to 100 mM NaCl over a period of 4 hours, and the first point taken immediately thereafter. Although stomatal conductance decreased markedly for several days, it did not result in limited carbon fixation. 

The success of treatment is measured by the early termination of seizures, without adverse drug effects or brain injury. Therefore, it is essential to start pharmacologic treatment as soon as possible. First-line treatment for SE should halt seizure activity within minutes of administration. In patients who are unarous-able following treatment, an EEG should be done to rule out continued excessive electrical brain activity and confirm termination of seizures. A physical exam and evaluation of the patient s laboratory results can help determine if the cause or complications of seizure activity are being appropriately treated. 

Although increases in bone mineral density have been reported at other sites, most of the clinically significant fractures occur in the hip or spine, and these sites have become clinically important measures in the trials. These increases in bone mineral density are an important marker of treatment effects and are related to the benefits found in larger trials of decreased fracture risk. 

Psychotherapy looks even better when its long-term effectiveness is assessed.17 Formerly depressed patients are far more likely to relapse and become depressed again after treatment with antidepressants than they are after psychotherapy. As a result, psychotherapy is significantly more effective than medication when measured some time after treatment has ended, and the more time that has passed since the end of treatment, the larger the difference between drugs and psychotherapy. This long-term advantage of psychotherapy over medication is independent of the severity of the depression. Psychotherapy outperforms antidepressants for severely depressed patients as much as it does for those who are mildly or moderately depressed.18... 

The effectiveness of pancreatic enzyme supplementation is measured by improvement in body weight and stool consistency or frequency. The 72-hour stool test for fecal fat may be used when the adequacy of treatment is in question. 

The measurement of the ethoxyresorufin-O-deethylase (EROD) activity is another sensitive parameter to detect the effects of paper mill industrial effluents on living organisms in the receiving waters. The EROD activity is a measure of the activity of the cytochrome P-450 enzyme system, which plays a central role in the transformation and elimination of xenobiotics. Increased EROD activity has been shown as far as 40 km from pulp mills, and EROD induction in fish caused by pulp mill effluents remains after biological treatment [60]. It is specified that EROD activity and erythrocytic nuclear abnormalities are induced by abietic and dehydroabietic acid [60]. 

---