目的 探讨住院急性痛风患者发热的危险因素。 方法 对符合纳入标准的185例住院急性痛风患者资料进行回顾性分析，根据体温分为中高热组（n=25）、低热组（n=25）和无发热组（n=135例）。比较3组临床资料、炎症指标及用药的资料差异。再根据有无膝关节疼痛分为膝关节疼痛组（n=69）和无膝关节疼痛组（n=116）比较2组炎症指标及穿刺治疗差异。采用二元logistic回归分析住院痛风发热的危险因素。 结果 3组在白细胞计数（9.2±2.3 vs. 8.6±3.3 vs. 7.7±2.5，P=0.022）、中性粒细胞百分比（77.7±4.0 vs. 70.5±10.6 vs. 67.7±12.1，P=0.001）、疼痛VAS评分［3（3.0，4.0） vs. 2（2.0，3.0） vs. 2（1.5，3.0），P=0.001］、C反应蛋白（c-reactive protein，CRP）［102（49，137） vs. 36（22，83） vs. 15（5，53），P=0.001］及血沉（erythrocyte sedimentation rate，ESR）［50（41，66） vs. 28（16，45） vs. 27（14，46），P=0.001］水平比较差异均有统计学意义，其中高热组在白细胞计数、中性粒细胞百分比比无发热组高（P=0.012、P=0.001）；中高热组在疼痛VAS评分、CRP、ESR比无发热组及低热组高（P=0.001、P=0.001、P=0.001及P=0.014、P=0.033、P=0.011）。低热组使用非甾体抗炎药的比例高于无发热组（92% vs 68.1%，P=0.015）及中高热组（92% vs. 60.0%，P=0.008）。3组在膝关节疼痛（56.0% vs. 48.0% vs. 31.9%，P=0.036）、第一足趾关节疼痛（6.0% vs. 0.0% vs. 25.9%，P=0.007）、上肢关节疼痛（40.0% vs. 16.0% vs. 17.8%，P=0.034）及多关节受累（56.0% vs. 24.0% vs. 25.9%，P=0.008）的比例比较差异均有统计学意义。有膝关节疼痛者ESR［43（21，56） vs. 25（14，41），P=0.001］、CRP［45（11，115） vs. 17（5，49），P=0.001］和关节腔局部治疗的比例（58.0% vs. 6.9%，P=0.001）较无膝关节受累组高。logistic回归分析发现白细胞（OR=1.171，95%CI=1.037~1.323，P=0.011）、中性粒细胞百分比（OR=1.053，95%CI=1.020~1.087，P=0.002）、CRP（OR=1.015，95%CI=1.008~1.021，P=0.001）、血沉（OR=1.023，95%CI=1.008~1.039，P=0.003）、疼痛VAS评分（OR=1.674，95%CI=1.228~2.282，P=0.001）、膝关节疼痛（OR=2.428，95%CI=1.252~4.709，P=0.009）为住院痛风发热的危险因素；第一足趾关节疼痛（OR=0.233，95%CI=0.068~0.804，P=0.021）为住院痛风发热的保护因素。其中CRP水平升高（OR=1.011，95%CI=1.005~1.018，P=0.001）和疼痛VAS评分的升高（OR=1.446，95%CI=1.035~2.019，P=0.031）为住院痛风发热的独立危险因素。 结论 住院急性痛风患者出现发热的比例为27%（50/185），发热患者有更高的炎症水平及疼痛评分，出现膝关节疼痛和多关节受累疼痛的比例更高，第一足趾关节疼痛的比例更低，需要更强的抗炎止痛治疗；尤其是CRP水平高，疼痛VAS评分高为住院痛风发热的独立危险因素。

Objective To investigate the risk factors of fever in hospitalized patients with acute gout. Methods The data of 185 hospitalized patients with acute gout meeting the inclusion criteria were retrospectively analyzed and divided into moderate and high fever group（n=25），low fever group（n=25） and no fever group（n=135） according to body temperature. The clinical data，inflammatory indicators and medication data were compared. Then the two groups were divided into knee pain group（n=69） and pain free group（n=116） according to the presence or absence of knee pain.The risk factors of hospitalized gout fever were analyzed by binary Logistic regression. Results The white blood cell count in 3 groups was（9.2±2.3 vs. 8.6±3.3 vs. 7.7±2.5，P=0.022），neutrophil percentage（77.7±4.0 vs. 70.5±10.6 vs. 67.7±12.1，P=0.001），pain VAS score［3（3.0，4.0） vs. 2（2.0，3.0） vs. 2（1.5，3.0），P=0.001］，c-reactive protein（102（49，137） vs. 36（22，83） vs. 15（5，53），P=0.001） and ESR levels50（41，66） vs. 28（16，45） vs. 27（14，46），P=0.001］ were statistically significant. The white blood cell count and neutrophil percentage in the high fever group were higher than those in the no-fever group（P=0.012，P=0.001）. The pain VAS score，CRP and ESR in moderate and high fever group were higher than those in no fever group and low fever group（P=0.001，P=0.001，P=0.001 and P=0.014，P=0.033，P=0.011）.The proportion of NSAIDS used in low-fever group was higher than that in no-fever group（92% vs. 68.1%，P=0.015） and medium-high fever group（92% vs. 60.0%，P=0.008）. In the three groups，knee pain（56.0% vs 48.0% vs. 31.9%，P=0.036），first toe joint pain（6.0% vs. 0.0% vs. 25.9%，P=0.007），upper limb joint pain（40.0% vs. 16.0% vs. 17.8%，P=0.034） and multiple joint involvement（56.0% vs. 24.0% vs. 25.9%，P=0.008） were statistically significant. Proportion of ESR［43（21，56） vs. 25（14，41），P=0.001］，CRP［45（11，115） vs. 17（5，49），P=0.001］ and local treatment of joint cavity in patients with knee pain［（58.0% vs. 6.9%），P=0.001］ were higher than those without knee joint involvement. Logistic regression analysis showed that leukocytes（OR=1.171，95%CI=1.037-1.323，P=0.011），neutrophil percentage（OR=1.053，95%CI=1.020-1.087，P=0.002），CRP（OR=1.015，95%CI=1.008~1.021，P=0.001），ESR（OR=1.023，95%CI=1.008-1.039，P=0.003），pain VAS score（OR=1.674，95%CI=1.228-2.282，P=0.001），knee pain（2.428，95%CI=1.252-4.709，P=0.009） was a risk factor for hospitalized gout fever. First toe joint pain（OR=0.233，95%CI=0.068-0.804，P=0.021） was a protective factor for hospitalized gout fever. The increase of CRP level（OR=1.011，95%CI=1.005-1.018，P=0.001） and pain VAS score（OR=1.446，95%CI=1.035-2.019，P=0.031） were independent risk factors for hospitalized gout fever. Conclusion The proportion of hospitalized acute gout patients with fever was 27%（50/185），and patients with fever had higher inflammation levels and pain scores，a higher proportion of knee pain and multi-joint pain，and lower rates of first toe joint pain，which required stronger anti-inflammatory and analgesic therapy. In particular，high CRP level and high pain VAS score were independent risk factors for hospitalized gout fever.