Thuraisingam Sharmala, Marasinghe D Himasara, Barrick Kendra, Aghajafari Fariba, Manski-Nankervis Jo-Anne, Dowsey Michelle M, Quan Hude, Williamson Tyler, Garies Stephanie
Department of Surgery, University of Melbourne, Melbourne, Australia.
Department of Family Medicine, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada.
J Med Internet Res. 2025 Jul 3;27:e69631. doi: 10.2196/69631.
General practice electronic health records (EHRs) contain a wealth of patient information. However, these data are collected for clinical purposes. Hence, questions remain around the suitability of using these data for other purposes, including epidemiological research, developing and validating clinical prediction models, conducting audits, and informing policy.
This study aimed to compare the quality of osteoarthritis-related data in Australian and Canadian general practice EHRs for externally validating a clinical prediction model for total knee replacement surgery.
A data quality assessment was conducted on 201,462 patient general practice EHRs from Australia provided by National Prescribing Service MedicineWise, and 92,425 from Canada provided by the Canadian Primary Care Sentinel Surveillance Network. Completeness, plausibility, and external validity of data elements relevant to osteoarthritis were assessed. Completeness and plausibility were evaluated using counts and proportions. For external validity, prevalence was estimated using proportions, and knee replacement summarized as a rate per 100,000 population.
There were minimal incomplete and implausible data fields for age and sex (<1%), geographic location (<5%), and commonly cooccurring comorbidities (<10%) in both datasets. However, weight, height, BMI, and Canadian Index of Multiple Deprivation contained >50% missing data. The recording of osteoarthritis by age and sex in both datasets were similar to national estimates, except for patients aged >80 years (Australia: 16.6%, 95% CI 16%-17.3% vs 13.1%, 95% CI 11.2%-15.4%; Canada: 36.7%, 95% CI 36.1%-37.2% vs 50.8%, 95% CI 50.7%-50.9%). Total knee replacement rates were substantially lower in both EHR datasets compared with national estimates (Australia: 72 vs 218 per 100,000; Canada: 0.84 vs 200 per 100,000).
Age, sex, geographic location, commonly cooccurring comorbidities, and prescribing of osteoarthritis medications in Australian and Canadian general practice EHRs are suitable for use in clinical prediction model validation studies. However, BMI and the Canadian Index of Multiple Deprivation are unfit for such use due to large proportions of missing data. Rates of total knee replacement surgery were substantially underreported and should not be used for prediction model validation. Better harmonization of patient data across primary and tertiary care is required to improve the suitability of these data. In the meantime, data linkage with national registries and other health datasets may overcome some of the data quality challenges in general practice EHRs.
全科医疗电子健康记录（EHRs）包含大量患者信息。然而，这些数据是为临床目的收集的。因此，围绕将这些数据用于其他目的（包括流行病学研究、开发和验证临床预测模型、进行审计以及为政策提供信息）的适用性仍存在问题。
本研究旨在比较澳大利亚和加拿大全科医疗EHRs中骨关节炎相关数据的质量，以对全膝关节置换手术的临床预测模型进行外部验证。
对澳大利亚国家处方服务机构MedicineWise提供的201,462份患者全科医疗EHRs以及加拿大初级保健哨点监测网络提供的92,425份患者全科医疗EHRs进行数据质量评估。评估与骨关节炎相关的数据元素的完整性、合理性和外部有效性。使用计数和比例评估完整性和合理性。对于外部有效性，使用比例估计患病率，全膝关节置换以每10万人口的发生率进行汇总。
两个数据集中，年龄、性别（<1%）、地理位置（<5%）以及常见合并症（<10%）的不完整和不合理数据字段极少。然而，体重、身高、BMI以及加拿大多重贫困指数缺失数据超过50%。两个数据集中按年龄和性别记录的骨关节炎情况与全国估计值相似，但80岁以上患者除外（澳大利亚：16.6%，95%CI 16%-17.3% 对比13.1%，95%CI 11.2%-15.4%；加拿大：36.7%，95%CI 36.1%-37.2% 对比50.8%，95%CI 50.7%-50.9%）。与全国估计值相比，两个EHR数据集中全膝关节置换率均显著较低（澳大利亚：每10万人中72例对比218例；加拿大：每10万人中0.84例对比200例）。
澳大利亚和加拿大全科医疗EHRs中的年龄、性别、地理位置、常见合并症以及骨关节炎药物处方适用于临床预测模型验证研究。然而，由于大量数据缺失，BMI和加拿大多重贫困指数不适合用于此类研究。全膝关节置换手术率报告严重不足，不应将其用于预测模型验证。需要更好地协调初级和三级医疗中的患者数据，以提高这些数据的适用性。同时，将数据与国家登记处及其他健康数据集相链接可能会克服全科医疗EHRs中的一些数据质量挑战。
