Journal of epidemiology and biostatistics最新文献

Background: Both small and large variant Creutzfeldt Jakob disease (vCJD) epidemics are consistent with the current observed incidence. Uncertainty in vCJD projections could potentially be reduced by incorporating information on the prevalence of the infectious agent in persons incubating vCJD. The prospect of vCJD prevalence studies has been raised by detection of abnormal prion protein, thought to be the infectious agent, in appendices and tonsils removed from vCJD patients. Although unlinked anonymous testing of stored operative tissues for abnormal prion protein is very appealing, the design and interpretation of such prevalence studies is complicated by the lack of information on how early in the incubation period of vCJD the abnormal prion protein becomes detectable.

Methods: We simulate a range of vCJD epidemics, consistent with the limited available information on the incidence of vCJD, to illustrate some of the potential problems encountered when interpreting the results from prevalence studies of detectable abnormal prion protein. We assume plausible incubation period distributions and dietary exposure patterns.

Results: We demonstrate, in the context of our simulated epidemics, that prevalence studies of detectable abnormal prion protein would require the testing of tens of thousands of operative specimens and, even then, that unlinked anonymous testing positives would be unexpected.

Background: Many epidemiological studies rely in part on proxy informants. There is little published information on the reliability of proxy-respondent reports of physical activity.

Methods: Self-reported data on vigorous and moderate physical activity, from a representative sub-sample of participants in a community-based case-control study of coronary heart disease, were compared with information collected from their next-of-kin.

Results: Relative to primary respondents, proxy respondents under-reported activity by approximately 10 percentage points, for both leisure and work-time activity. On a simple three point scale (inactivity/moderate activity/physical activity), 70% of primary-proxy pairs were in exact agreement with regard to leisure time activity and 67% of pairs were in exact agreement on work-time activity. The corresponding values for the weighted kappa statistic were 0.66 [95% confidence interval (CI) 0.59-0.72] and 0.62 (0.54-0.72). Log-linear modelling provided evidence for superior agreement on worktime activity when the proxy was not the primary respondent's spouse.

Discussion: Overall levels of primary-proxy respondent agreement on physical activity seem somewhat lower than has been reported for smoking and alcohol-drinking frequency. There seems little reason to prefer spouse proxies when endeavouring to elicit information on work-time physical activity. Log-linear modelling provides an efficient means of exploring covariate effects in observer-agreement studies.

Background: A food-frequency questionnaire (FFQ) used to assess usual food intake in Southern France has been validated by the classical means of multiple-day food records. To minimise over-estimation of the correlation between the dietary assessments by the FFQ and the reference method, which occurs if the random errors of questionnaire and reference measurement are positively correlated, a triangular comparison, the method of triads, was used.

Methods: We applied the triads model by comparing the FFQ with two multiple-day food records and three biomarkers. Only 87 subjects were included and completed the protocol. One biomarker (beta-carotene) was used for the 87 subjects and two biomarkers (urinary nitrogen and potassium) were measured in only 40 subjects.

Results: For beta-carotene intake assessment, the triad model, including the weighed multiple records (PETRA), was the best with estimates of validity coefficient of 0.39 [confidence interval (CI) 0.18-0.60] for the FFQ, 0.52 (CI 0.24-0.86), for PETRA and 0.85 (CI 0.43-1) for plasma levels of the nutrient. For protein and potassium intake assessment, the triad model including the estimated multiple records was the best only for the estimates of FFQ validity coefficient (0.61; CI 0.28-0.96 and 0.31; CI 0.09-0.66 respectively).

Conclusion: Accuracy of the dietary assessment methods permitted a satisfactory estimation of the validity coefficient for beta-carotene intake by the FFQ, despite a small sample. However, the validity coefficients for protein and potassium showed wide CI values, indicating that a sample size < 50 subjects appears unsatisfactory for validation.

Background: Responses to questions on sensitive behaviours can be affected by when and how questions are asked.

Methods: Using data from an ongoing survey of women taking isotretinoin (a teratogenic drug used to treat acne), we compared answers to questions on contraceptive use during therapy among 126966 women who had begun treatment in 1990-93. 20503 were randomly assigned to a group which was surveyed by telephone at the start of treatment, during treatment and 6 months after treatment ended (DAT group). 106463 received a mail survey 6 months after treatment ended (AT group). Women who could not be contacted during treatment, along with the non-respondents in the AT group, received follow-up surveys by US mail, air courier and, if necessary, were called by telephone.

Results: The response rate was higher in the DAT group (96.4%) than in the AT group (84.8%). Among respondents, the odds of reported contraceptive use in the AT group compared with the DAT group (asked at 6 months after treatment) was 1.41 [95% confidence interval (CI): 1.36-1.46]. However, rates of oral contraceptive (OC) use were virtually identical in the two groups. In both groups, women surveyed by telephone reported lower rates of contraceptive use than women surveyed by mail. In the AT group, a late response was associated with a significantly lower rate of contraceptive use.

Conclusions: Both mode of data collection (phone versus mail) and time to response affected reported rates of contraceptive use. Reasonable assumptions about the effect of non-response could not account for the difference between the AT and DAT groups. We found that contacting women during their treatment did not increase their reported use of contraceptives.

Background: A method was developed for stochastically reconstructing the pattern of infection from observed epidemic data. This allowed for estimation of a time-dependent force of infection, or transmission rate, during an epidemic.

Methods: A discrete-time mechanistic model was used to describe the spread of infection and a stochastic procedure, which utilised the latent and infectious period distributions, was used to reconstruct the dates of infection, becoming infectious and removal from the given data. The four equations describing the model were then solved to obtain least squares estimates of the transmission rate and the basic reproduction number (R0) throughout the epidemic. This process was repeated in order to assess the variability in these key epidemiological parameters. The stochastic epidemic reconstruction procedure was developed to account for changes in the distribution of the survival period over the course of the epidemic and adapted for application to epidemic data where not all infected individuals have yet been observed as cases.

Results: The method was applied to a set of epidemic data from an outbreak of classical swine fever in Pakistan. Constant and time-varying estimates of the transmission rate were derived and compared. There was some evidence to suggest that the force of infection varied over time.

Discussion: The method described can be applied to data from epidemics where observations are incomplete. The confidence limits obtained for the estimated force of infection provide a means of assessing the evidence for time variation in this parameter.

Background: The Scandinavian countries, Denmark, Norway and Sweden, have established both HIV and AIDS registers to monitor the HIV epidemic. Information in such registers can be used to estimate the number of new HIV infections over time, incidence rates and prevalence. Information from the HIV registers made it possible to study what kind of effects such information had in the estimation process, compared with using information about new AIDS cases only.

Methods: A Markov model back-calculation approach was used. One model incorporated data on cases of both HIV and AIDS. Another model incorporated data on cases of AIDS only. Death or emigration prior to the onset of AIDS and effects of treatment were included in both models.

Results: Estimates of absolute rates of HIV for men who have sex with men (MSM) showed a distinct development in each country. Significant differences in incidence rates and prevalence of HIV among MSM were found between Scandinavian countries when information on diagnosed HIV was incorporated. Precision was improved when using both HIV and AIDS diagnosed cases compared with using AIDS cases only. The epidemic in Denmark was more extensive than in the two other countries for the whole study period.

Discussion: The results were fairly robust against reasonable variation in the model parameters. The more extensive epidemic in Denmark may have been caused by the homosexual culture denying that HIV was a disease more relevant to them than to others, until the HIV test was publicly available in 1985.

Background: Recent reports indicated that previous infection may be a risk factor for ischaemic stroke in younger adults and that the increased mortality of cerebrovascular diseases in winter months may be partly caused by the increased rate of infection during the cold season.

Methods: We performed a 1:1 matched case-control study with 197 cases (83 females, 114 males) aged between 22 and 80 years (median age 65 years) to investigate risk factors for acute cerebrovascular ischaemia, in particular the effect of previous infection. We estimated the impact of risk factors in terms of attributable and absolute risks.

Results: All risk factors together, previous infection, hypertension, diabetes mellitus, smoking, coronary heart disease, previous stroke or transient ischaemic attack and family history of stroke yield a summary attributable risk of 0.74 [95% confidence intervals (CI) 0.64-0.83]. Recent infections showed a relative risk of 4.3 (95% CI 1.8-10.5) and an attributable risk of 0.15 (95% CI 0.09-0.21). Seventeen percent of the German population are estimated to be in a high-risk group. This subgroup contributes about 55% of the estimated yearly 120,000 incident cases in the age group 55-84 in Germany.

Discussion: Identification of high-risk groups for stroke on the basis of individual risk factor distribution and the estimation of its size is possible and may produce useful results. Reducing the prevalence of infection and early treatment of bacterial infection may lower the incidence of stroke.

Background: Age-period-cohort analyses of US breast-cancer mortality rates reveal an unexpected decrease in risk for women born after 1948. Hormones are thought to play an important role in the aetiology of breast cancer and female gynaecologic cancers, and thus the evaluation of birth-cohort trends for female gynaecologic cancers may shed light on the declining breast-cancer risk among 'baby-boomers'.

Methods: Age-period-cohort analyses are applied to US mortality rates for breast cancer, ovarian cancer, endometrial cancer and cervical cancer from 1950 through 1995.

Results: Age-period-cohort analyses provide no clues regarding the declining birth-cohort risk for breast cancer in 'baby-boomers'. The birth-cohort curves for ovarian and endometrial cancers are roughly similar, and largely explained by known risk factors. The calendar-period curve for endometrial cancer reveals increased risk between 1960 and 1980, probably due to increased use of oestrogen replacement therapy. Changes in the birth-cohort curve for cervical cancer reflect, for the most part, changes in sexual activity. An unexpected significant increase in the calendar-period curve for ovarian cancer occurred around 1980.

Conclusion: Most of the major changes in the calendar-period and birth-cohort curves for breast cancer and female gynaecologic cancers can be explained by documented changes in known risk factors and in medical practice. The decreasing breast-cancer birth-cohort risk among 'baby-boomers' and the increasing ovarian-cancer calendar-period curve after 1980 are recent changes that require further investigation.