Cancer Epidemiology and Services Research, Sydney School of Public Health, University of Sydney, Sydney, NSW, Australia.
Surgical Outcomes Research Centre, Sydney Local Health District and University of Sydney, Sydney, NSW, Australia.
Med J Aust. 2014 Apr 21;200(7):403-7. doi: 10.5694/mja13.10710.
To identify predictors of variation in colorectal cancer care and outcomes in New South Wales.
DESIGN, SETTING AND PATIENTS: Multilevel logistic regression analysis using a linked population-based dataset based on the records of patients with cancer of the colon, rectosigmoid junction or rectum who were registered in 2007 and 2008 by the NSW Central Cancer Registry and treated in 105 hospitals in NSW.
Six outcome measures (30-day mortality, 28-day emergency readmission, prolonged length of stay, 30-day wound infection, 90-day venous thromboembolism, 1-year mortality) and five care process measures (discussion at multidisciplinary team [MDT] meeting, documented cancer stage, recorded pathological stage, treatment within 31 days of decision to treat, treatment within 62 days of referral).
We analysed data for 6890 people. There was wide variation between hospitals in care process measures, even after adjusting for patient and hospital factors. Older adults were less likely to be discussed at an MDT meeting and receive treatment within suggested time frames (all P < 0.001 for colon cancer). Increasing patient age, greater extent of disease, higher Charlson comorbidity score and resection after emergency admission consistently showed strong evidence of an association with poor outcomes. Much of the variation between hospitals in outcome measures was accounted for by patient characteristics.
Patient characteristics should be included in risk-adjustment models for comparing outcomes between hospitals and for quantifying hospital variation. Further exploration of the reasons why certain hospitals and patients appear to be at risk of poorer care is needed.
确定新南威尔士州结直肠癌治疗和结局变化的预测因素。
设计、设置和患者：使用基于人群的链接数据集进行多水平逻辑回归分析，该数据集基于 2007 年和 2008 年新南威尔士州中央癌症登记处登记的结肠癌、直肠乙状结肠交界处或直肠癌症患者的记录，并在新南威尔士州 105 家医院接受治疗。
六个结局指标（30 天死亡率、28 天急诊再入院、住院时间延长、30 天伤口感染、90 天静脉血栓栓塞、1 年死亡率）和五个治疗过程指标（多学科团队[MDT]会议讨论、记录癌症分期、记录病理分期、在决定治疗后 31 天内进行治疗、在转诊后 62 天内进行治疗）。
我们分析了 6890 人的数据。即使在调整了患者和医院因素后，医院之间的治疗过程指标也存在很大差异。老年人不太可能在 MDT 会议上讨论，也不太可能在建议的时间框架内接受治疗（所有结肠癌患者 P < 0.001）。患者年龄增加、疾病程度增加、Charlson 合并症评分增加以及急诊后行切除术与不良结局有很强的关联。医院之间在结局指标上的大部分差异都归因于患者特征。
在比较医院之间的结局和量化医院之间的差异时，应将患者特征纳入风险调整模型。需要进一步探讨为什么某些医院和患者似乎面临较差治疗的风险的原因。
