Discussion
A new prediction tool for the risks of tophi formation amongst gout patients was developed and validated in this study. The nomogram consisted of nine easily available variables, including hospitalisation frequency, disease duration, number of joints involved in gouty arthritis, gout flares frequency, smoking, and whether combined with atherosclerosis, diabetes, hypertension and kidney dysfunction. Through the analysis using C-index and calibration plot alongside internal verification, the nomogram was verified to have good discrimination and calibration ability, and could be widely used in gout patients to relatively accurately predict the risks of tophi formation.22
Based on the results of predictor analysis, the nomogram demonstrated that hospitalisation for more than twice per year, disease duration of more than ten years, more than three joints involved in gouty arthritis, frequency of gout flares for more than twice per year, smoking, and combined with atherosclerosis, diabetes, hypertension and kidney dysfunction could be the predictive factors for tophi formation. Amongst the nine risk factors, most were consistent with the results of previous studies.2,23 A retrospective clinical study involving 5,693 Chinese gout patients depicted a significant difference (P<0.05) between the tophi and non-tophi groups with regard to disease duration (10.28±7.54 years vs. 5.11±6.06 years), number of joints involved (3.11±2.15 vs. 1.81±1.35), systolic pressure (138.53±19.46 mmHg vs. 133.87±17.93 mmHg), diastolic pressure (89.55±12.73 mmHg vs. 87.48±11.77 mmHg) and creatinine clearance rate (Ccr) (93.05±48.7 mL/min vs. 106.61±51.76 mL/ min).2This study’s results were similar to those findings, which might be explained by certain reasons. First of all, tophi is one of the manifestations of chronic gout as a deposition of uric acid crystals.24 Similar to other chronic metabolic diseases, a longer gout duration entails more severe conditions, leading to more joints involved in gouty arthritis, more frequent gout flare and hospitalisation, as well as a higher probability of tophi formation. Secondly, previous studies have discussed that uric acid can increase blood pressure through damaging the vascular endothelium, activating the renin angiotensin aldosterone system, and reducing the nitric oxide levels in endothelium.25 Meanwhile, hypertension can damage the renal blood perfusion and lead to tissue hypoxia, thus reducing the uric acid clearance rate. Therefore, a more severe gout leads to a higher risk of hypertension.26 Likewise, a more serious hypertension denotes a more likelihood to cause kidney damage and a less excreted uric acid that entails a higher risk of tophi formation.27 In addition, previous researches have indicated that smokers and diabetic patients have higher uric acid concentrations, which can lead to vascular endothelial dysfunction, thus increasing the risk of atherosclerosis, renal dysfunction and tophi formation.28-30
Based on the aforementioned risk factors, it is of great significance to enable certain measures to prevent and control tophi formation at an early stage. According to the nomogram, for patients with longer gout duration, apart from paying attention to controlling the blood uric acid level and reducing gout flare frequency as well as the number of joints involved in gouty arthritis, we should also highly consider monitoring blood glucose, blood pressure, renal function and vascular conditions in order to reduce the risks of tophi formation. For patients who already have the aforementioned risk factors, early diagnosis and treatment of tophi is necessary to avoid significant tophi-related complications such as irreversible joint deformity and dysfunction, bone destruction and fracture, infection and ulceration.
A series of steps had been undertaken to minimise potential bias in this study. Firstly, for patients who have been hospitalised more than once, only the data from the first hospitalisation were included to avoid inclusion of repeated cases. In the electronic medical record system, patients diagnosed with ‘gout’ or ‘gouty arthritis’ were searched using keywords or ICD codes to avoid missing patients, which might cause selection bias. Finally, uncertain data that were recorded in the hospital system were reconfirmed by contacting the patients to ensure data accuracy.
Despite the above measures, there are still some limitations in the current study. First of all, the sample size was relatively small; most subjects were male, and came from Zhejiang Province, China. Additionally, the sample size of the tophi group was small, which might limit the power of the regression analysis. Secondly, the predictive model might not have included all potential factors related to tophi formation. Since all data were collected from medical records of discharged patients during hospitalization, some variables such as serum uric acid and body mass index, might not represent patients’ actual uric acid and weight control. Besides, certain demographics such as family history and smoking or drinking history might be inconsistent with the actual situation, due to patients’ intentional concealment or unintentional false memories. Thirdly, despite the efforts to examine the robustness of the nomogram by internal verification using a bootstrap test, it is still possible that the nomogram is not quite useful for gout populations in other regions and countries. Hence, an external validation in wider populations is necessary in future studies.