How many patients are required

Although in the vast majority of cancer trials it will be most appropriate to have duration of survival as the primary outcome measure and perform power calculations for this outcome measure, it is nevertheless important and useful to calculate a required sample size for the QL aspect.

For example, the primary aim of the QL aspect of a lung cancer trial might be to compare the difference in the proportion of patients who have relief from their cough at three months. If the standard treatment is expected to give 50 per cent of patients relief, then to reliably detect an increase in this proportion to 75 per cent with the new treatment would require a total of 154 patients (seventy-seven per arm).

However, this sample size assumes that all patients will be assessable. In many advanced cancers, attrition due to death, even at three months, may have reduced the trial population by 10 or 15 per cent. In addition, the above scenario requires complete information on all patients and indeed all patients to start with a cough otherwise how could relief be assessed? The framing of the hypothesis is again important, as, for example, Stephens etal. [43] have suggested redefining 'palliation' as not just improvement, but also prevention (for asymptomatic patients), and control (for patients with minimal symptoms). In this way all patients could be included in the analysis. Nevertheless, the calculated QL sample size must be adjusted by adding in a realistic proportion of patients to account for deaths, ineligibility, and non-compliance. For instance, in a multicentre trial looking at palliation in poor performance non-small cell lung cancer patients it maybe prudent to increase the sample substantially. Obviously this number depends on the timing of the primary QL outcome, the expected attrition and likely compliance.

There may be situations within a large trial where it is only necessary that a subset of patients complete QL questionnaires. This may be because the trial is powered to detect a small difference in duration of survival, but a larger difference in QL. Any decision to assess only a subset of patients must be based on carefully worked out hypotheses and sample size, and it is usually important to ensure that the patients in the subset are representative of the whole group (if you wish to draw conclusions relevant to the whole group), although occasionally a specific subset of patients may be selected for study. However, generally there are a number of options that could be applied when deciding how to acquire a subset sample.

♦ Collect QL data on all patients until the subset sample size is achieved. This is probably the best option. It should ensure no bias in the sample, assuming that over time the characteristics of patients entering the trial remain approximately the same, and that there is no substantial learning curve in the management of patients.

♦ Limit the collection of QL data to a few centres. Although this may ensure good compliance (one would naturally choose the centres with most enthusiasm for QL), it is often extremely difficult to predict the number of patients who will be entered from each centre, or whether good compliance can be maintained if, for instance, a key staff member in the local centre leaves. However, it is unlikely that a subset based on centres would be unrepresentative of the whole population, in the way that, for example, a subset based on performance status would be.

♦ Have QL as an option either for centres or patients. This is the least attractive option. Giving patients the option will almost inevitably result in an unrepresentative sample, as those patients who choose to complete QL may well be the youngest and/or fittest. Experience shows that giving centres the option will almost inevitably cause problems, as many centres, given an option to reduce their administrative burden, will opt out.

A Disquistion On The Evils Of Using Tobacco

A Disquistion On The Evils Of Using Tobacco

Among the evils which a vitiated appetite has fastened upon mankind, those that arise from the use of Tobacco hold a prominent place, and call loudly for reform. We pity the poor Chinese, who stupifies body and mind with opium, and the wretched Hindoo, who is under a similar slavery to his favorite plant, the Betel but we present the humiliating spectacle of an enlightened and christian nation, wasting annually more than twenty-five millions of dollars, and destroying the health and the lives of thousands, by a practice not at all less degrading than that of the Chinese or Hindoo.

Get My Free Ebook

Post a comment