Conversely, correlations in errors between individuals, groups, or time represent shared error. Measurement error that is assumed to be independent and identically distributed is said to be unshared. Dose estimation error comprises both systematic and random components, where systematic error represents an inequality between the long-term averages of true and observed dose to individual, i, in group, j, at time, t, and random errors represent natural variation in X ij( t) and Z ij( t). Additional information is provided in Appendix A and a companion paper ( 27). Table 1 describes dose estimation errors considered in this review. All dosimetrists convened to discuss, reconcile, and consolidate disparate findings within a category to achieve consensus. At least two dosimetrists independently reviewed each study within a category. Investigators categorized studies as environmental, occupational, or medical exposure. When studies stemmed from the same population and observation period, selection was based on consensus of the monograph working group ( 7). When more than one article pertained to a study population, the synthesis was limited to the study with the longest follow-up. Articles from the Fifteen-Country Workers Study ( 9) published within the eligible period were excluded because main findings were reviewed in BEIR VII ( 2). Exposures stemmed mainly from external gamma and x-rays however, contributions from other sources (eg, incorporated radionuclides or neutron exposures) were considered in some studies. The primary exposure was to sparsely ionizing radiation at low doses and low-dose rates resulting in an average absorbed dose to the whole-body or target tissue of interest of 100 mGy or less. Studies were either cohort or case-control designs with dose-response analyses of the relationship between cumulative radiation dose and cancer, reporting effect measures in terms of risk per unit dose or at a given exposure level. Briefly, investigators systematically searched public domain databases for epidemiologic studies on radiation-exposed populations published from 2006 through 2017. As a part of this assessment, the work herein is a systematic appraisal of the strengths and weaknesses of the dosimetry systems used in these studies, including an assessment of sources and effects of potential discrepancies between the true absorbed dose to target tissues from ionizing radiation (ie, the preferred dose quantity) and the value used in dose-response analyses, hereafter referred to as dose estimation error.ĭetails on study selection are provided elsewhere ( 7). The National Cancer Institute assessment largely followed recent guidance by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) for evaluating radiation epidemiologic studies ( 8). The National Cancer Institute is leading an effort by international experts to critically evaluate a group of post–BEIR VII studies and assess several potential sources of biases on estimates of risk from low-dose ionizing radiation exposure ( 7). The relevant literature has grown considerably since BEIR VII. The review examined a wide array of information from medically, occupationally, and environmentally exposed populations however, the Committee again relied on LSS data to estimate risk because of uncertainty in risk estimates and a general lack of accounting for errors in dose estimation in other studies.
In 2006, the National Research Council of the National Academy of Sciences’ Committee on the Biological Effects of Ionizing Radiation (BEIR) published its most recent review of existing data on health effects from low levels of ionizing radiation, hereafter referred to as BEIR VII ( 2). The direct estimation of risk is preferred when data are sufficient, and health risks from ionizing radiation in several populations have been studied extensively ( 2, 3). The acute ionizing radiation exposure in the LSS population differs from the protracted lower dose rate exposures in most occupational and environmental settings therefore, the transport of risk in the LSS to other populations (eg, radiation workers) is uncertain ( 4–6). Since the 1950s, authoritative bodies have relied mostly on data from the Life Span Study (LSS) of Japanese atomic bomb survivors to project cancer risks from ionizing radiation exposure ( 2, 3). The foremost concern about low-dose ionizing radiation exposure is the potential for increased risk of cancer ( 1). Ionizing radiation exposure is unavoidable in everyday life.
0 kommentar(er)
