Chrzanowski Jędrzej, Michalak Arkadiusz, Łosiewicz Aleksandra, Kuśmierczyk Hanna, Mianowska Beata, Szadkowska Agnieszka, Fendler Wojciech
Department of Biostatistics and Translational Medicine, Diabetology, Endocrinology and Nephrology, Medical University of Lodz, Lodz, Poland.
Department of Department of Pediatrics, Diabetology, Endocrinology and Nephrology, Medical University of Lodz, Lodz, Poland.
Diabetes Technol Ther. 2021 Apr;23(4):293-305. doi: 10.1089/dia.2020.0433. Epub 2021 Mar 9.
Accurate estimation of glycated hemoglobin (HbA1c) from continuous glucose monitoring (CGM) remains challenging in clinic. We propose two statistical models and validate them in real-life conditions against the current standard, glucose management indicator (GMI). Modeling utilized routinely collected data from patients with type 1 diabetes from central Poland (eligibility criteria: age >1 year, diabetes duration >3 months, and CGM use between 01/01/2015 and 12/31/2019). CGM records were extracted from dedicated Medtronic/Abbott databases and cross-referenced with HbA1c values; 28-day periods preceding HbA1c measurement with >75% of the sensor-active time were analyzed. We developed a mixed linear regression, including glycemic variability indices and patient's ID (glucose variability-based patient specific model, GV-PS) intended for closed-group use and linear regression using patient-specific error of GMI (proportional error-based patient agnostic model, PE-PA) for general use. Models were validated with either new HbA1cs from closed-group patients or separate patient-HbA1c pool. External validation was performed with data from clinical trials. Performance metrics included bias, its 95% confidence interval (95% CI), coefficient of determination (), and root mean square error (RMSE). We included 723 HbA1c-CGM pairs from 174 patients (mean age 9.9 ± 4.4 years and diabetes duration 3.7 ± 3.6 years). GMI yielded = 0.58, with different bias between Medtronic and Abbott devices [0.120% vs. -0.152%, < 0.0001], and overall 95% CI = -0.9% to +1%, RMSE = 0.47%. GV-PS successfully captured patient-specific variance (closed-group validation: = 0.83, bias = 0.026%, 95% CI = -0.562% to 0.591%, RMSE = 0.31%). PE-PA performed similarly on new patients ( = 0.76, bias = -0.069%, 95% CI = -0.790% to 0.653%, RMSE = 0.37%). In external validation GMI, GV-PS, and PE-PA produced 73.8%, 87.5%, and 91.0% predictions within 0.5% (5.5 mmol/mol) from the true value. Constructed models performed better than GMI. PE-PA provided an accurate estimate of HbA1c with fast and straightforward implementation.
在临床中，通过连续血糖监测（CGM）准确估算糖化血红蛋白（HbA1c）仍然具有挑战性。我们提出了两种统计模型，并在实际生活条件下针对当前标准血糖管理指标（GMI）对其进行了验证。建模利用了从波兰中部1型糖尿病患者中常规收集的数据（纳入标准：年龄>1岁，糖尿病病程>3个月，且在2015年1月1日至2019年12月31日期间使用CGM）。CGM记录从专用的美敦力/雅培数据库中提取，并与HbA1c值进行交叉对照；分析了HbA1c测量前28天内传感器活跃时间>75%的时间段。我们开发了一种混合线性回归模型，包括血糖变异性指标和患者ID（基于血糖变异性的患者特异性模型，GV - PS），用于封闭组使用，以及使用GMI的患者特异性误差的线性回归模型（基于比例误差的患者通用模型，PE - PA）用于一般使用。模型通过来自封闭组患者的新HbA1c或单独的患者 - HbA1c池进行验证。使用临床试验数据进行外部验证。性能指标包括偏差及其95%置信区间（95%CI）、决定系数（）和均方根误差（RMSE）。我们纳入了174例患者的723对HbA1c - CGM数据（平均年龄9.9±4.4岁，糖尿病病程3.7±3.6年）。GMI的=0.58，美敦力和雅培设备之间存在不同的偏差[0.120%对 - 0.152%，<0.0001]，总体95%CI = - 0.9%至 + 1%，RMSE = 0.47%。GV - PS成功捕捉了患者特异性差异（封闭组验证：=0.83，偏差=0.026%，95%CI = - 0.562%至0.591%，RMSE = 0.31%）。PE - PA在新患者中表现类似（=0.76，偏差= - 0.069%，95%CI = - 0.790%至0.653%，RMSE = 0.37%）。在外部验证中，GMI、GV - PS和PE - PA在与真实值相差0.5%（5.5 mmol/mol）范围内的预测分别为73.8%、87.5%和91.0%。构建的模型比GMI表现更好。PE - PA能够快速、直接地实施，提供准确的HbA1c估算值。
