Silverstein M D, Ballard D J
Department of Health Sciences Research, Mayo Clinic, Rochester, USA.
J Health Serv Res Policy. 1998 Jul;3(3):134-40. doi: 10.1177/135581969800300303.
To compare an expert panel's global assessment of appropriateness of elective surgery for abdominal aortic aneurysms (AAA) with their assessment of the effect of surgery on the probability of 5-year mortality.
Nine expert panel members rated the appropriateness of 120 scenarios for elective AAA repair on a nine-point scale, and also estimated the 5-year probability of AAA-related death and of non-AAA related death among 30-day survivors of AAA surgery and among patients with unoperated AAA. These probabilities were used to determine differences in 5-year probability of mortality of surgery vs. no surgery for each scenario. Three categories of appropriateness were defined based on these differences: inappropriate (< 0%), equivocal (0-5%), and appropriate (> 5%).
The distribution of scenarios was inappropriate (39%), equivocal (12%), and appropriate (49%) based on probability estimates and inappropriate (43%), equivocal (22%), and appropriate (36%) based on global assessment. There was poor agreement between the two methods, with a Kappa coefficient = 0.28 (95% CI: 0.23 to 0.32). Although a higher proportion of scenarios were rated as appropriate using probability estimation rather than global judgment, the level of agreement among members of the panel was similar, Kappa coefficient = 0.07 (95% CI: -0.07 to 0.72).
Experts disagree about the appropriate indications for elective surgery for AAA. Explicit estimates used in a decision analysis may provide a better assessment of appropriate indications than the global judgment of experts. Global assessment of the appropriateness of AAA surgery based on panel members' review of research evidence for increased survival appears to include implicitly their valuation of outcomes.
比较一个专家小组对腹主动脉瘤（AAA）择期手术适宜性的整体评估，以及他们对手术对5年死亡率概率影响的评估。
九名专家小组成员对120个腹主动脉瘤择期修复方案的适宜性进行九点量表评分，并估计腹主动脉瘤手术30天幸存者以及未接受手术的腹主动脉瘤患者中与腹主动脉瘤相关死亡和非腹主动脉瘤相关死亡的5年概率。这些概率用于确定每种情况下手术与非手术5年死亡率的差异。根据这些差异定义了三类适宜性：不适宜（<0%）、模棱两可（0 - 5%）和适宜（>5%）。
根据概率估计，方案分布为不适宜（39%）、模棱两可（12%）和适宜（49%）；根据整体评估，不适宜（43%）、模棱两可（22%）和适宜（36%）。两种方法之间的一致性较差，Kappa系数 = 0.28（95%CI：0.23至0.32）。尽管使用概率估计而非整体判断时，被评为适宜的方案比例更高，但专家小组成员之间的一致程度相似，Kappa系数 = 0.07（95%CI： - 0.07至0.72）。
专家们对腹主动脉瘤择期手术的适宜指征存在分歧。决策分析中使用的明确估计可能比专家的整体判断能更好地评估适宜指征。基于小组成员对生存获益研究证据的审查对腹主动脉瘤手术适宜性进行的整体评估似乎隐含了他们对结果的评估。
