Gill T M, Horwitz R I
Department of Medicine, Yale University School of Medicine, New Haven, CT 06520-8025.
J Clin Epidemiol. 1995 Feb;48(2):281-92. doi: 10.1016/0895-4356(94)00129-e.
To evaluate the efficacy of cancer screening, investigators have turned increasingly to the case-control approach as an efficient alternative design to the randomized controlled trial. Overlooked in the design and implementation of screening case-control studies, however, have been important clinical distinctions that are crucial for decisions in patient management and health policy. These distinctions involve the classification of baseline risk factors and prognostic co-morbidity, the enumeration and definition of the intervention, and the choice and timing of the outcome. When investigators neglect the distinctive characteristics of patients and their diseases, screening case-control studies may have reduced validity and impaired utility. To strengthen the validity of these studies, investigators can analyze co-morbidity suitably, report a "clinical" confidence interval as the estimate of screening efficacy, carefully evaluate the attribution of death, and allow sufficient follow-up time for the death of patients diagnosed with cancer. Moreover, to strengthen the generalizability of screening case-control studies, investigators can stratify their findings for pertinent baseline risk factors and can expand the screening definition. If implemented, these guidelines may enhance the capacity of case-control studies to provide clinicians and policy-makers with critically needed information on the efficacy of cancer screening.
为评估癌症筛查的效果，研究人员越来越多地采用病例对照研究方法，将其作为随机对照试验的一种有效替代设计。然而，在筛查病例对照研究的设计和实施过程中，一些重要的临床差异被忽视了，而这些差异对于患者管理决策和卫生政策至关重要。这些差异涉及基线风险因素和预后合并症的分类、干预措施的列举和定义，以及结局的选择和时机。当研究人员忽视患者及其疾病的独特特征时，筛查病例对照研究的有效性可能会降低，实用性也会受损。为提高这些研究的有效性，研究人员可以适当地分析合并症，报告一个“临床”置信区间作为筛查效果的估计值，仔细评估死亡原因，并为癌症确诊患者的死亡留出足够的随访时间。此外，为增强筛查病例对照研究的可推广性，研究人员可以根据相关基线风险因素对研究结果进行分层，并扩大筛查定义。如果实施这些指导原则，可能会提高病例对照研究为临床医生和政策制定者提供有关癌症筛查效果急需信息的能力。
