Stevens Lauren A S, Spangler Leslie, Yochum Laura, Ding Yan, Wang Florence T
Optum Epidemiology, Optum, Boston, MA, USA.
Center for Observational Research, Amgen, Inc., Thousand Oaks, CA, USA.
Clin Epidemiol. 2021 Oct 28;13:1019-1026. doi: 10.2147/CLEP.S325080. eCollection 2021.
Medication-related osteonecrosis of jaw (MRONJ) is associated with certain drug therapies. Pharmacoepidemiologic studies often rely on electronic healthcare data to assess adverse events following drug exposure. Few studies have developed and validated claims-based MRONJ identification algorithms. This study assessed the performance of claims-based MRONJ algorithms by chart review of potential cases among postmenopausal (PM) women and women with postmenopausal osteoporosis (PMO).
Among PM and PMO women sourced from a large US commercial health insurance database affiliated with Optum, potential cases were identified by International Classification of Diseases, 9th and 10th Revisions (ICD-9, ICD-10) diagnosis codes; 200 were selected for chart retrieval, with the goal of obtaining 100 charts in each coding era. Procured charts were redacted and then reviewed by an oral surgeon who determined case status. Positive predictive values (PPV) and 95% confidence intervals (CI) were calculated overall, by cohorts, and coding eras. Baseline characteristics were assessed. Two potential algorithm refinements were explored: using a restricted set of ICD codes; requiring antibiotic use after MRONJ diagnosis.
A total of 1273 potential cases were identified. Of the 200 potential cases selected, 104 (52%) were procured, and six cases were confirmed (PPV 5.8%, 95% CI 2.2, 12.1). Baseline characteristics were largely similar across all strata. Potential algorithm refinements yielded marginal PPV improvement.
This study identified a small number of confirmed cases, and the resulting PPVs were low, but consistent with reported studies. Potential algorithm refinements yielded minimal improvements. To our knowledge, this study is the first to report on the identification of MRONJ using ICD-10 codes in the US.
药物相关性颌骨坏死（MRONJ）与某些药物治疗相关。药物流行病学研究通常依靠电子医疗数据来评估药物暴露后的不良事件。很少有研究开发并验证基于索赔的MRONJ识别算法。本研究通过对绝经后（PM）女性和绝经后骨质疏松症（PMO）女性中的潜在病例进行病历审查，评估基于索赔的MRONJ算法的性能。
从与Optum相关的大型美国商业健康保险数据库中选取PM和PMO女性，通过国际疾病分类第9版和第10版（ICD - 9、ICD - 10）诊断代码识别潜在病例；选择200例进行病历检索，目标是在每个编码时代获取100份病历。获取的病历进行编辑后由口腔外科医生进行审查以确定病例状态。总体、按队列和编码时代计算阳性预测值（PPV）和95%置信区间（CI）。评估基线特征。探索了两种潜在的算法改进：使用一组受限的ICD代码；要求在MRONJ诊断后使用抗生素。
共识别出1273例潜在病例。在选取的200例潜在病例中，获取了104例（52%），确诊6例（PPV 5.8%，95% CI 2.2，12.1）。所有分层的基线特征基本相似。潜在的算法改进使PPV有小幅提高。
本研究识别出少量确诊病例，所得PPV较低，但与已报道的研究一致。潜在的算法改进效果甚微。据我们所知，本研究是美国首次报告使用ICD - 10代码识别MRONJ的研究。
