Validation of the Women's Health Initiative Insomnia Rating Scale in a multicenter controlled clinical trial.

OBJECTIVE: The objective of this study was to evaluate the construct validity of the five-item Women's Health Initiative Insomnia Rating Scale (WHIIRS) by comparing women taking hormone therapy (HT) versus those taking a placebo and by comparing women known to differ in vasomotor symptoms. METHODS: The WHIIRS was included in two phase III randomized trials intended to evaluate the efficacy of a combination estradiol plus and norethindrone acetate transdermal delivery system in reducing vasomotor symptoms. In all, 850 healthy postmenopausal women participated in these studies. Both trials were double-blind, one was placebo-controlled and the other was positive-controlled. The former trial admitted women with > or =8 hot flashes/day and lasted 12 weeks with data collected on the WHIIRS at baseline, 4, 8, and 12 weeks. The other trial had no entry criteria pertaining to hot flashes and lasted 52 weeks with WHIIRS data collected at baseline, 12, 24, and 52 weeks. RESULTS: The WHIIRS was sensitive to the effect of HT on sleep disturbance over time. The WHIIRS also detected differences in self-reported sleep disturbance between women with mild vasomotor symptoms compared with those with moderate to severe symptoms. As expected, the study using a positive control revealed that sleep improved over time (p <.0001). Also as predicted, the study using a placebo control found that sleep disturbance in the treatment groups improved at a faster rate than in the control groups (p = .035). CONCLUSION: The construct validity of the WHIIRS was supported because it was successfully used to detect self-reported sleep disturbance differences in women taking HT versus those taking a placebo as well as in groups known to differ in severity of their vasomotor symptoms.

Theorizing about causes at the individual level while estimating effects at the population level: implications for prevention.

The dominant philosophy of modern epidemiology is individualism, despite the limitations of epidemiologic tools and methods when considering the individual level. We pursue information on increasingly reductionist causes in our search for knowledge of causes of specific cases. Philosophical reasoning and empiric evidence suggest that this search may not be as fruitful as proponents claim. I argue that using individualism to guide our search for causes of disease hinders our effectiveness in contributing to disease prevention, because the positive predictive values of most established genetic and environmental risk factors for noninfectious diseases are too low to be quantitatively convincing to an individual.

Validation of the Harvard Cancer Risk Index: a prediction tool for individual cancer risk.

OBJECTIVE: Risk appraisal tools are increasingly being used in the clinical setting to estimate individuals' risks of developing and dying from diseases. The Harvard Cancer Risk Index is one such tool constructed to predict the risks of individuals, aged 40 and above, for developing the leading types of cancer in U.S. men and women relative to the general population. To date, the Risk Index has not been prospectively validated. STUDY DESIGN AND SETTING: Over a period of 10 years' follow-up in the Nurses' Health Study and the Health Professionals' Follow-up Study, age-standardized incidence ratios for cancer of the ovary, colon, and pancreas were calculated for the Risk Index's relative risk categories to assess goodness of fit for risk prediction at the aggregate level. Age-adjusted concordance statistics were determined as measures of discriminatory accuracy at the individual level. RESULTS: The Risk Index was well calibrated with observed relative risks across categories for ovarian and colon cancer in women and pancreatic cancer in men, while it performed moderately for colon cancer in men. Discriminatory accuracy was modest for ovarian cancer (age-adjusted concordance statistic = 0.59), and relatively good for pancreatic cancer (concordance statistic of 0.72), and colon cancer in men and women (concordance statistics of 0.71, 0.67 respectively). CONCLUSION: The results of this prospective validation provide evidence for the validity of the Risk Index in predicting individuals' risks of cancers, and thereby offer support for future applications of this risk appraisal tool.

Breast cancer risk prediction with a log-incidence model: evaluation of accuracy.

OBJECTIVE: We examined whether a breast cancer risk prediction model other than the Gail et al. model performs better at discriminating between women who will and who will not develop the disease. METHODS: We applied the two published versions of the Rosner and Colditz log-incidence model of breast cancer, developed on data from the Nurses' Health Study, to the estimation of 5-year risk for the period 1992 to 1997 in the same cohort. The first version contained reproductive factors only, and the second version contained a more extensive list of risk factors. RESULTS: Both versions of the model fit well. The ratio of expected to observed numbers of cases (E/O) in the first version was 1.00 (95% confidence interval [CI] 0.93-1.07); for the extended version the E/O was 1.01 (95% CI 0.94-1.09). The age-adjusted concordance statistic was 0.57 for the first model version and 0.63 for the extended version. CONCLUSION: The discriminatory accuracy of the two versions was modest, although the addition of the variables in the extended version meaningfully increased the discriminatory accuracy of risk prediction over that found with the more parsimonious model.