Background older patients with hip fractures are arbitrarily classified as octogenarians, nonagenarians and centenarians. We have designed this study to quantify in-hospital mortality and complications among each of these groups. We hypothesised that the associations between age and in-hospital mortality and complications are continuously increasing, and that these risks increase rapidly when patients reach a certain age. Methods this research is a retrospective cohort study using nationwide database between 2010 and 2018. Patients undergoing hip fracture surgery, and aged 60 or older, were included. The associations between patient age, in-hospital mortality and complications were visualised using the restricted cubic spline models, and were analysed employing multivariable regression models. Then, octogenarians, nonagenarians and centenarians were compared. Results among a total of 565,950 patients, 48.7% (n = 275,775) were octogenarians, 23.0% (n = 129,937) were nonagenarians and 0.7% (n = 4,093) were centenarians. The models presented three types of association between age, in-hospital mortality and complications: (i) a continuous increase (mortality and respiratory complications); (ii) a mild increase followed by a steep rise (intensive care unit admission, heart failure, renal failure and surgical site hematoma) and (iii) a steep increase followed by a limited change (coronary heart disease, stroke and pulmonary embolisms). Conclusion we identified three types of association between age and clinical outcomes. Patients aged 85–90 may constitute the upper threshold for age categorisations, because the risk of in-hospital complications changed dramatically at that stage. This information can improve clinical awareness of various complications and support collective decision-making.

中文翻译：

---

髋部骨折手术后的年龄和临床结果：八十多岁、九十多岁和百岁老人分类重要吗？

背景 患有髋部骨折的老年患者被任意分为八十多岁、九十多岁和百岁老人。我们设计了这项研究来量化这些组中的每一个的住院死亡率和并发症。我们假设年龄与住院死亡率和并发症之间的关联不断增加，并且当患者达到一定年龄时，这些风险会迅速增加。方法本研究为2010-2018年全国数据库的回顾性队列研究。纳入60岁及以上接受髋部骨折手术的患者。使用限制三次样条模型可视化患者年龄、住院死亡率和并发症之间的关联，并使用多变量回归模型进行分析。那么，八十多岁的人，比较了九岁和百岁老人。结果在总共 565,950 名患者中，48.7% (n = 275,775) 为 80 多岁，23.0% (n = 129,937) 为 90 岁，0.7% (n = 4,093) 为百岁老人。这些模型展示了年龄、住院死亡率和并发症之间的三种关联：(i) 持续增加（死亡率和呼吸系统并发症）；(ii) 轻微上升后急剧上升（重症监护病房入院、心力衰竭、肾衰竭和手术部位血肿）和 (iii) 急剧上升后出现有限变化（冠心病、中风和肺栓塞）。结论 我们确定了年龄和临床结果之间的三种关联。85-90 岁的患者可能构成年龄分类的上限，因为在那个阶段，院内并发症的风险发生了巨大变化。这些信息可以提高对各种并发症的临床认识并支持集体决策。