An overview of systematic reviews investigating the accuracy, reliability, and relationships of tests and measures for diagnosis of neck pain.

Background: Neck pain is a common condition and often difficult to differentiate and diagnose. Previous literature has investigated diagnostic accuracy of examination measures but is limited. Objective: To investigate examination measures for diagnosis of neck pain. Design: Umbrella review Literature Search: Four databases were searched electronically. Quality and risk of bias were assessed using the AMSTAR 2 and ROBIS. Diagnostic criteria were investigated for accuracy, reliability, and relationship to neck pain diagnoses. Study Selection Criteria: Systematic reviews of randomized clinical trials evaluating diagnostic criteria for neck pain. Data Synthesis: Twenty seven systematic reviews were included. Quantitative and qualitative results were summarized in narrative format. Results: Hand radiculopathy and numbness have good specificities (0.89-0.92) for facet and uncinate joint hypertrophy. The extension rotation test (ERT) and manual assessment have good sensitivities and moderate-good specificities. Positive ERT combined with positive manual assessment findings (+LR = 4.71; Sp = 0.83) improves diagnostic accuracy compared to a positive ERT alone (+LR = 2.01; Sp = 0.59). Canadian C-spine Rules and Nexus low-risk criteria have excellent validity in screening for cervical fracture or instability. Imaging appears to have validity in diagnosing ligamentous disruption or fractures, but lacks clarity on predicting future neck pain. Increased fatty infiltrates have been found with whiplash associated disorders and mechanical neck pain. Conclusions: There are limited indicators providing strong diagnostic value for cervical spine diagnosis. Strength of recommendations are limited by heterogeneous outcomes, methodology, and classification systems. Future research should attempt to provide stronger recommendations of differential diagnostic criteria for pathoanatomical neck pain diagnoses.

研究背景：颈痛是一种常见临床病症，其鉴别与诊断往往存在较大难度。既往已有文献针对各类临床检查手段的诊断准确性展开相关探索，但现有研究成果较为有限。
研究目的：探讨适用于颈痛诊断的各类检查手段。
研究设计：伞状综述（Umbrella review）。
文献检索：通过电子检索方式覆盖4个数据库。采用AMSTAR 2与ROBIS工具对纳入研究的质量及偏倚风险进行评估。针对颈痛诊断相关标准的准确性、可靠性及其与颈痛诊断的相关性展开分析。
研究筛选标准：纳入针对颈痛诊断标准进行评估的随机对照试验所开展的系统综述。
数据综合：最终纳入27项系统综述研究，以叙述性综述形式对其中的定量与定性研究结果进行汇总分析。
研究结果：手部神经根病与麻木症状对于关节突及钩椎关节肥大具有良好的特异性（0.89~0.92）。伸旋转试验（Extension Rotation Test, ERT）与手法评估具备良好的敏感度，且特异性处于中等至良好水平。单独阳性ERT结果的阳性似然比（+LR）为2.01，特异性（Sp）为0.59；将阳性ERT与阳性手法评估结果相结合时，阳性似然比可达4.71，特异性为0.83，诊断准确性较单独阳性ERT结果显著提升。加拿大颈椎规则（Canadian C-spine Rules）与Nexus低风险标准在筛查颈椎骨折或颈椎失稳方面具备极佳的效度。影像学检查在诊断韧带撕裂或骨折方面具备一定效度，但在预测未来颈痛发生的应用中仍缺乏明确依据。在鞭打相关疾患与机械性颈痛患者中，可观察到脂肪浸润程度升高的现象。
研究结论：目前能够为颈椎诊断提供较强诊断价值的临床指标较为有限。推荐意见的强度受研究结局异质性、研究方法学差异以及分类系统不统一的限制。未来研究应致力于为病理解剖学层面的颈痛诊断提供更具说服力的鉴别诊断标准推荐方案。