本研究旨在调查在布里格姆和妇女医院以及丹纳·法伯癌症研究所（DFCI）对血液病理学中的算法测试的有效性。该算法基于初始病理评估后的测试选择，以最大程度提高费效比测试，尤其针对昂贵的分子和细胞遗传学检测。为17种疾病类别开发了标准测试（SOPs）订单方案，并编码在决策支持应用程序中。获得了为期六个月的应用程序β测试的回顾性数据，并与该时间段的实际测试实践进行了比较。此外，还从一个社区分店获得了2年的前瞻性数据。总共分析了460个回顾性病例（算法测试引入之前）和109个前瞻性病例（引入后）。 在回顾性数据中，61.7％的测试（825项中的509项）与SOPs一致，而有38.3％的测试（825项中的316项）过度订购，有30.8％（736项中的227项）的SOP推荐测试被省略。在前瞻性数据中，98.8％的测试（247项中的244项）是一致的，只有1.2％过度订购（247项中的3项），并且有7.6％的测试被省略（264个SOP推荐测试中的20项；整体P <0.001）。在布里格姆和妇女医院/DFCI，SOP实施之前过度订购测试的成本显示出潜在的年化节省费用为$1,347,520美元。只有2个过度订购测试（0.6％）提供了额外的信息，且仅适用于极为罕见的临床情况。SOP的实施大大改善了测试订购实践，具有适量的辅助测试，可以最大程度减少成本，并对获得关键信息提供结果没有显著影响。

This study investigated the effectiveness of algorithmic testing in hematopathology at the Brigham and Women's Hospital and Dana-Farber Cancer Institute (DFCI). The algorithm was predicated on test selection after an initial pathologic evaluation to maximize cost-effective testing, especially for expensive molecular and cytogenetic assays.Standard ordering protocols (SOPs) for 17 disease categories were developed and encoded in a decision support application. Six months of retrospective data from application beta testing was obtained and compared with actual testing practices during that timeframe. In addition, 2 years of prospective data were also obtained from patients at one community satellite site.A total of 460 retrospective cases (before introduction of algorithmic testing) and 109 prospective cases (following introduction) were analyzed. In the retrospective data, 61.7% of tests (509 of 825) were concordant with the SOPs while 38.3% (316 of 825) were overordered and 30.8% (227 of 736) of SOP-recommended tests were omitted. In the prospective data, 98.8% of testing was concordant (244 of 247 total tests) with only 1.2% overordered tests (3 of 247) and 7.6% omitted tests (20 of 264 SOP-recommended tests; overall P < .001). The cost of overordered tests before implementing SOP indicates a potential annualized saving of $1,347,520 in US dollars (USD) in overordered testing at Brigham and Women's Hospital/DFCI. Only two of 316 overordered tests (0.6%) returned any additional information, both for extremely rare clinical circumstances.Implementation of SOPs dramatically improved test ordering practices, with a just right number of ancillary tests that minimizes cost and has no significant impact on acquiring key informative test results.