Correcting overall survival for the impact of crossover via a rank-preserving structural failure time (RPSFT) model in the RECORD-1 trial of everolimus in metastatic renal-cell carcinoma.

Clinical trials in oncology often allow patients randomized to placebo to cross over to the active treatment arm after disease progression, leading to underestimation of the treatment effect on overall survival as per the intention-to-treat analysis. We illustrate the statistical aspects and practical use of the rank-preserving structural failure time (RPSFT) model with the Fleming-Harrington family of tests to estimate the crossover-corrected treatment effect, and to assess its sensitivity to various weighting schemes in the RECORD-1 trial. The results suggest that the benefit demonstrated in progression-free survival is likely to translate into a robust overall survival benefit.

The use of simulation in the design of two cardiovascular survival studies.

Simulation has become practical as an everyday tool given the wide availability of high performance workstations. We argue that simulation can play an important role in determining the number of patients required in a survival study, particularly if the assumption of proportional hazards does not hold, or if the study design is complex. The argument is illustrated by two examples. The first considers the design of a post myocardial infarction survival trial in which the hazard ratio is not constant. The second provides sample size estimates for a multicentre heart failure study in which both the treatment effect and the control mortality rate vary across centres.

Sample sizes for proportional hazards survival studies with arbitrary patient entry and loss to follow-up distributions.

In proportional hazards survival studies, power depends on the observed number of deaths, d*. For a given choice of survival, loss, and patient entry distributions, sample sizes can be determined by equating d* to the expected number of deaths. Approximating the survival and loss distributions with piecewise exponential distributions, and patient entry with a piecewise linear distribution, significantly reduces the computational overhead, and the expected number of deaths can be evaluated routinely. The merits of this approach are illustrated by a clinical trial of chemotherapy for large bowel cancer.

Demonstration of insulin transformation products in insulin vials by high-performance liquid chromatography.

OBJECTIVE: To assess the importance of temperature, light, agitation, and regular withdrawal of insulin on the rates of appearance of three groups of insulin transformation products (ITPs) in vials of human and beef soluble insulin. RESEARCH DESIGN AND METHODS: Twelve insulin-treated patients (6 receiving human soluble insulin, 6 receiving beef soluble insulin) participated in the study. Reverse-phase high-performance liquid chromatography was used to measure ITP in serial samples from vials stored in various laboratory environments or issued to patients. Vials in the laboratory were analyzed on nine occasions over 26 wk; those issued to patients were analyzed weekly. RESULTS: Results were expressed as rates of change (% total protein/wk). Except for human insulin stored at 4 degrees C, sampling did not accelerate transformation. Storage at higher temperatures caused more rapid transformation to all ITP groups, and human insulin was more susceptible than beef insulin. Exposure to light accelerated transformation mainly to group 3 ITP. Fibrillation occurred in beef but not human insulin carried in a shirt pocket. Results from patients' vials were consistent with reported storage conditions. CONCLUSIONS: All three ITP groups form in human and beef soluble insulin. Their rate of production is a function of temperature and light exposure. Differences between the two insulins were probably due to differences in formulation. Patients should be discouraged from hoarding vials and storing insulin in direct sunlight. One or more ITPs could contribute to the residual immunogenicity of modern insulin preparations.

Employment problems and diabetes.

A survey of employment problems in a random sample of diabetic patients and a group of control subjects aged 17-65 years was carried out in eight centres in the UK. Data were linked to information collected from patients' diabetic clinic notes relating to the presence and treatment of any diabetic complications and quality of diabetic control. Difficulties in obtaining employment because of diabetes were reported by 13% of diabetic patients, and because of illness by 2% of control subjects (p less than 0.001). Nine percent of diabetic patients and 2% of control subjects reported having to change their job because of their illness (p less than 0.001), and 7% of people with diabetes and 2% of people without diabetes reported losing a job because of their illness (p less than 0.001). Diabetic shift workers were twice as likely as control subjects working shifts to experience problems with their job (18 vs 8%, p = 0.045). Reports of any sickness absence in the last 12 months were not significantly different for people with and without diabetes (49 vs 45%). Sickness absence in excess of 20 days in the last 12 months was more common among diabetic patients than control subjects (29 vs 16%, p less than 0.001). People with diabetes are more likely to experience problems in obtaining employment and staying employed than people without diabetes.

Unemployment and diabetes.

Unemployment was examined in a random sample of diabetic clinic attenders and a group of non-diabetic control subjects aged 17-65 years, selected from eight different centres in Great Britain. Data on employment were obtained using a postal questionnaire and medical details were collected from diabetic clinic notes. Out of a 61% response rate, 22% of males and 12% of females with diabetes were unemployed compared with 8 and 5% of the control group (both p less than 0.001). A greater percentage of people with diabetes were economically inactive, that is retired, unable to work, ill or housewives compared with the control group (29 vs 14%, p less than 0.001). Young people with diabetes (16-25 years) had the highest rates of unemployment. A matched pairs analysis confirmed that diabetic men had higher unemployment rates than their controls (14 vs 7%, p less than 0.001). Comparisons were made between unemployment rates for the eight geographical areas and published unemployment statistics; unemployment rates were significantly higher for men with diabetes except at one centre. A stepwise multiple logistic regression analysis indicated that variables which were predictive of unemployment were similar to those expected for people without diabetes. Unemployment is apparently a problem for the person with diabetes, especially for the young.

Employers' attitudes to diabetes.

In order to examine the knowledge and attitude of employers to diabetes and diabetic employees, a questionnaire was sent to a random sample of approximately 2500 businesses and industries in eight different areas of the United Kingdom. Information was collected on the employment of people with diabetes, such as sickness absence, suitability of types of job for the diabetic person, the proportion of people with diabetes registered as disabled, and the provision of pension schemes. Only 1% of firms claimed that they would not consider a person with diabetes for employment within the firm. Nearly a third of employers did not know whether they employed anyone with diabetes. Over a quarter of the manufacturing industries reported some jobs that were unsuitable and 8% reported certain unfavourable job conditions for a diabetic employee such as shift work, working at heights, strenuous work, etc. A total of 7% of those firms with a diabetic employee reported that diabetic workers were more often absent from work due to sickness and 16% of firms employing a diabetic person did not allow paid time off for clinic visits. The attitude of the majority of employers seemed to indicate that diabetes in itself does not limit employment prospects since most people with diabetes have few problems arising from their condition and can make good employees in a range of occupations.

Psychosocial factors and the onset of type 1 diabetes.

Stress has been proposed as a possible precipitating factor for the development of Type 1 diabetes in genetically susceptible individuals. A validated in-depth psychometric instrument (The Life Events and Difficulties Schedule) was used to investigate the role of psychosocial factors in the onset of Type 1 diabetes in islet cell antibody (ICA)-positive family members. Families with ICA-positive members had higher scores for subject focused severe life events in the 5-year period prior to the diagnosis of diabetes in a second family member compared with matched control families (1.50 +/- 0.68 (+/- SD) vs 0.32 +/- 0.37; p less than 0.03). In the same time period, these families also experienced a higher mean number of severe long-term difficulties compared with control families (1.34 +/- 0.52 vs 0.14 +/- 0.17; p less than 0.01). Scores for current total visual social contacts were greater for control families compared with case family members (12.10 +/- 1.90 vs 9.64 +/- 2.70; p less than 0.05). These results suggest that global family stress possibly in conjunction with a reduced number of social contacts may act as a trigger for the development of diabetes in a second family member and that social support may act as a buffer to stress and disease onset.