目的：评价便携式非接触式眼压计(PNCT)与传统非接触式眼压计(NCT)测量不同眼压水平的准确性。

方法：回顾性研究。纳入2025年1月至2026年1月在宁波市眼科医院就诊的患者。所有患者均完成PNCT、NCT和Goldmann压平眼压计(GAT)测量。根据GAT测量值分为低眼压组(GAT≤16 mmHg)、中等眼压组(16 mmHg结果：本研究共纳入患者94例184眼，其中男52例，女42例，平均年龄48.6±16.4岁，GAT、PNCT、NCT测量眼压值分别为 \〖19.3(15.5，24.0)、19.3(16.0，24.1)、19.0(15.0，26.0)mmHg\〗(χ² = 3.77， P>0.05)。Spearman相关分析显示，PNCT和NCT与GAT测量的眼压值均呈正相关(r_s=0.86、0.84，P<0.05)。ICC分析显示，PNCT与GAT的ICC=0.89(P<0.05)，NCT与GAT的ICC=0.88(P<0.05)，三组仪器整体ICC=0.88(P<0.05)。Bland-Altman分析显示，PNCT与GAT测量的眼压平均偏倚为0.1 mmHg(95%LoA：-7.3-7.6 mmHg)； NCT与GAT测量的眼压平均偏倚为0.1 mmHg(95%LoA：-8.4-8.6 mmHg)。根据GAT测量的眼压值进行分组，低眼压组60眼(32.6%)眼压≤16 mmHg，中等眼压组63眼(34.2%)16 mmHg<眼压<23 mmHg，高眼压组61眼(33.2%)眼压≥23 mmHg。低眼压组中PNCT和GAT测量值平均偏倚为1.5 mmHg(95%LoA：-4.4-7.4mmHg)； NCT和GAT测量的眼压平均偏倚为0.5 mmHg(95%LoA：-5.6-6.6 mmHg)； 中等眼压组中PNCT和GAT测量眼压平均偏倚为-0.2 mmHg(95%LoA：-5.4-5.1 mmHg)； NCT和GAT测量眼压平均偏倚为-0.1 mmHg(95%LoA：-8.6-8.4 mmHg)； 高眼压组中PNCT和GAT测量眼压平均偏倚为-0.8 mmHg(95%LoA：-10.6-8.9 mmHg)； NCT和GAT测量眼压平均偏倚为-0.1 mmHg(95%LoA：-10.5-10.4 mmHg)。本研究共有175眼完成了中央角膜厚度( CCT)测量，平均CCT为553.3±35.9(450-660)μm。GAT(r_s=0.31)、PNCT(r_s=0.39)及NCT(r_s=0.40)的测量值均与CCT呈正相关(P<0.05)。PNCT与GAT测量值的差值与CCT无相关性(r_s=0.08， P>0.05)； NCT与GAT测量值的差值与CCT呈正相关(r_s=0.23，P<0.05)。

结论：PNCT和NCT与GAT测量眼压结果相关性高、一致性良好，而且PNCT受CCT影响更小。但随着眼压升高，PNCT和NCT的测量误差也有所增大。PNCT和NCT的测量结果只能作为参考，均不能代替GAT。

METHODS:Patients who presented to Ningbo Eye Hospital between January 2025 and January 2026 were retrospectively included. All patients underwent measurements using PNCT, NCT, and Goldmann applanation tonometry(GAT). Based on the GAT measurements, the patients were categorized into a low IOP group(IOP≤16 mmHg), a moderate IOP group(16 mmHgRESULTS: A total of 94 patients(184 eyes)were included in this study, comprising 52 males and 42 females with a mean age of 48.6±16.4 y. The median IOP values measured by GAT, PNCT, and NCT were 19.3(15.5, 24.0), 19.3(16.0, 24.1), and 19.0(15.0, 26.0)mmHg, respectively(χ²=3.77, P>0.05). Spearman correlation analysis revealed that the IOP values measured by both PNCT and NCT were positively correlated with those measured by GAT(r_s=0.86 and 0.84, respectively, P<0.05). Intraclass correlation coefficient(ICC)analysis demonstrated an ICC of 0.89 between PNCT and GAT(P<0.05), and 0.88 between NCT and GAT(P<0.05), with an overall ICC of 0.88 among the three instruments(P<0.05). Bland-Altman analysis indicated a mean bias of 0.1 mmHg \〖95% limits of agreement(LoA): -7.3 to 7.6 mmHg\〗 between PNCT and GAT, and a mean bias of 0.1 mmHg(95%LoA: -8.4 to 8.6 mmHg)between NCT and GAT. Based on the GAT measurements, the eyes were categorized into a low IOP group(60 eyes, 32.6%, IOP≤16 mmHg), a moderate IOP group(63 eyes, 34.2%, 16 mmHgLoA: -4.4 to 7.4 mmHg)for PNCT and 0.5 mmHg(95%LoA: -5.6 to 6.6 mmHg)for NCT. In the moderate IOP group, the mean bias was -0.2 mmHg(95%LoA: -5.4 to 5.1 mmHg)for PNCT and -0.1 mmHg(95%LoA: -8.6 to 8.4 mmHg)for NCT. In the high IOP group, the mean bias was -0.8 mmHg(95%LoA: -10.6 to 8.9 mmHg)for PNCT and -0.1 mmHg(95%LoA: -10.5 to 10.4 mmHg)for NCT. Furthermore, 175 eyes underwent central corneal thickness(CCT)measurements, with a mean CCT of 553.3±35.9(450-660)μm. The IOP measurements from GAT(r_s=0.31), PNCT(r_s=0.39), and NCT(r_s=0.40)all exhibited positive correlations with CCT(P<0.05). Notably, the measurement difference between PNCT and GAT was not correlated with CCT(r_s=0.08, P>0.05), whereas the difference between NCT and GAT showed a positive correlation with CCT(r_s=0.23, P<0.05).

CONCLUSION:The IOP measurements obtained by PNCT, NCT, and GAT demonstrated strong correlation and good agreement, with PNCT being less influenced by CCT. However, as IOP levels increased, the measurement errors of both PNCT and NCT also increased. Therefore, PNCT and NCT readings should only serve as references and cannot replace GAT.