Centre for Clinical Epidemiology, Biostatistics and Health Services Research, the University of Melbourne, Australia.
Acad Emerg Med. 2012 Apr;19(4):439-47. doi: 10.1111/j.1553-2712.2012.01327.x.
The objective was to identify predictive factors and outcomes associated with patients who leave emergency departments (EDs) without being seen in Victoria, Australia.
This was a retrospective observational study of Victorian ED patient visits between July 1, 2000, and June 30, 2005, using linked hospital, ED, and death registration data. Index ED visits were identified for patients who left without being seen (LWBS) and for those who completed ED treatment and were discharged home. Statistical analyses included a general description and univariate analysis of patient, ED visit, temporal, and hospital-level factors. Logistic regression models were developed to assess risk factors associated with LWBS status compared to patients who completed treatment, to assess 48 hour re-presentations to ED; 48-hour hospital admissions; and 2-,7-, and 30-day mortality among those who LWBS compared to those who completed treatment. Adjusted odds ratios (ORs) and 99% confidence intervals (CIs) are presented.
There were 239,305 LWBS episodes, for 205,500 patients over the 5-year period. Independent factors associated with LWBS patients in comparison to those who completed treatment include patients who are younger (15 to 24 years, OR = 2.46, 99% CI = 2.37 to 2.56), male (OR = 1.07, 99% CI = 1.05 to 1.08), of Australian indigenous background (OR = 1.63, 99% CI = 1.53 to 1.73), of non-English-speaking background (OR = 1.08, 99% CI = 1.06 to 1.10), noncompensable status (OR = 1.73, 99% CI = 1.68 to 1.79), self-referring (OR = 1.46, 99% CI = 1.43 to 1.49), nonassisted arrival mode (OR = 1.35, 99% CI = 1.30 to 1.40), and those with a hospital admission in the 12 months before the ED presentation (OR = 1.53, 99% CI = 1.51 to 1.55). Patients who LWBS had triage categories of lower urgency (nonurgent, OR = 8.21, 99% CI = 8.00 to 8.43), attended during the evening (OR = 1.10, 99% CI = 1.08 to 1.12), on either Sunday (OR = 1.20, 99% CI = 1.18 to 1.23) or Monday (OR = 1.20, 99% CI = 1.17 to 1.23), in winter (OR = 1.14, 99% CI = 1.12 to 1.16), with higher rates occurring in higher volume EDs (OR = 2.20, 99% CI = 2.15 to 2.26). There was no greater risk of mortality for LWBS patients compared to patients who completed treatment. The risk of hospital admission within 48 hours of discharge was lower for LWBS patients (OR = 0.60, 99% CI = 0.58 to 0.62); however, ED re-presentation risk was higher (OR = 1.63, 99% CI = 1.60 to 1.67).
Patients who leave EDs in Victoria, Australia, without being seen are at lower risk of hospital admission and at no greater risk of mortality, but are at higher risk of re-presenting to an ED compared to patients who complete treatment and are discharged home.
本研究旨在识别与澳大利亚维多利亚州急诊科（ED）未就诊患者相关的预测因素和结局。
这是一项回顾性观察性研究，使用了维多利亚州 2000 年 7 月 1 日至 2005 年 6 月 30 日期间的医院、ED 和死亡登记数据，对患者进行了匹配。对未就诊（LWBS）和完成 ED 治疗并出院回家的患者进行了索引 ED 就诊的识别。统计分析包括患者、ED 就诊、时间和医院水平因素的一般描述和单变量分析。Logistic 回归模型用于评估与完成治疗的患者相比，LWBS 状态的相关风险因素，以评估 48 小时内再次到 ED 就诊；48 小时内入院；LWBS 患者与完成治疗患者相比，2、7 和 30 天的死亡率。报告了调整后的优势比（OR）和 99%置信区间（CI）。
在 5 年期间，有 239305 例 LWBS 发作，涉及 205500 名患者。与完成治疗的患者相比，LWBS 患者的独立因素包括年龄在 15 至 24 岁之间（OR=2.46，99%CI=2.37 至 2.56）、男性（OR=1.07，99%CI=1.05 至 1.08）、澳大利亚土著背景（OR=1.63，99%CI=1.53 至 1.73）、非英语背景（OR=1.08，99%CI=1.06 至 1.10）、非补偿状态（OR=1.73，99%CI=1.68 至 1.79）、自我转诊（OR=1.46，99%CI=1.43 至 1.49）、非辅助到达模式（OR=1.35，99%CI=1.30 至 1.40），以及在 ED 就诊前 12 个月内有医院入院记录的患者（OR=1.53，99%CI=1.51 至 1.55）。LWBS 患者的分诊类别较低（非紧急，OR=8.21，99%CI=8.00 至 8.43），在晚上就诊（OR=1.10，99%CI=1.08 至 1.12），在周日（OR=1.20，99%CI=1.18 至 1.23）或周一（OR=1.20，99%CI=1.17 至 1.23）就诊，在冬季就诊（OR=1.14，99%CI=1.12 至 1.16），较高的发生率发生在较高容量的 ED 中（OR=2.20，99%CI=2.15 至 2.26）。与完成治疗的患者相比，LWBS 患者的死亡率没有更高的风险。LWBS 患者出院后 48 小时内入院的风险较低（OR=0.60，99%CI=0.58 至 0.62）；然而，ED 再次就诊的风险较高（OR=1.63，99%CI=1.60 至 1.67）。
澳大利亚维多利亚州急诊科未就诊的患者入院和死亡的风险较低，但与完成治疗并出院回家的患者相比，再次到 ED 就诊的风险较高。
