オーストラリア全国肺がんスクリーニングプログラムの適格患者をプライマリケアで特定する：臨床意思決定支援システムを使用した横断研究およびPLCOm2012データ品質の評価

Primary care is pivotal in identifying eligible participants for Australia's National Lung Cancer Screening Program (NLCSP). Clinical Decision Support Systems (CDSS) can assist by flagging potentially eligible individuals using Electronic Medical Record (EMR) data. We developed a CDSS to identify people aged 50-70 who currently or formerly smoked cigarettes (with an unknown quit date or cessation within the past 10 years) for use in Australian general practice EMRs. This study aimed to: (1) assess the algorithm's accuracy in identifying eligible patients, (2) estimate NLCSP eligibility based on audited EMR smoking data, and (3) evaluate EMR data quality for applying the PLCO_m2012 risk prediction tool. A cross-sectional EMR audit was conducted across five Victorian general practices. Of 4,186 records flagged by the CDSS, 1,315 (31.4 %) were manually audited. The algorithm correctly identified all patients according to its rules (age and smoking status). Pack-year could be calculated for 226 patients (17.2 %), with 91 (6.9 %) meeting NLCSP eligibility. Eligibility was higher among people who currently smoke (9.2 %) than those who formerly smoked (3.6 %). A subset of 933 records was assessed for PLCOm2012 variables, revealing substantial data gaps (% missing): education level (100 %), ethnicity (58 %), BMI (57 %), and daily cigarette consumption (33 %). CDSS algorithms can identify potential NLCSP participants, but confirmation of smoking history remains essential due to sub-optimal smoking data quality. Significant data limitations currently hinder PLCO_m2012 implementation via CDSS in general practice EMRs.