Background

The prevalence of malnutrition in patients with cancer is reported as high as 60% to 80%, and malnutrition is associated with lower survival, reduced response to treatment, and poorer functional status. The Malnutrition Screening Tool (MST) is a validated tool when administered by health care professionals; however, it has not been evaluated for patient-led screening.

Objectives

This study aims to assess the reliability of patient-led MST screening through assessment of inter-rater reliability between patient-led and dietitian-researcher–led screening and intra-rater reliability between an initial and a repeat patient screening.

Design/participants

This cross-sectional study included 208 adults attending ambulatory cancer care services in a metropolitan teaching hospital in Queensland, Australia, in October 2016 (n=160 inter-rater reliability; n=48 intra-rater reliability measured in a separate sample).

Main outcome measurements

Primary outcome measures were MST risk categories (MST 0-1: not at risk, MST ≥2: at risk) as determined by screening completed by patients and a dietitian-researcher, patient test–retest screening, and patient acceptability.

Statistical analysis

Percent and chance-corrected agreement (Cohen’s kappa coefficient, κ) were used to determine agreement between patient-MST and dietitian-MST (inter-rater reliability) and MST completed by patient on admission to unit (patient-MSTA) and MST completed by patient 1 to 3 hours after completion of initial MST (patient-MSTB) (intra-rater reliability).

Results

High inter-rater reliability and intra-rater reliability were observed. Agreement between patient-MST and dietitian-MST was 96%, with “almost perfect” chance-adjusted agreement (κ=0.92, 95% CI 0.84 to 0.97). Agreement between repeated patient-MSTA and patient-MSTB was 94%, with “almost perfect” chance-adjusted agreement (κ=0.88, 95% CI 0.71 to 1.00). Based on dietitian-MST, 33% (n=53) of patients were identified as being at risk for malnutrition, and 40% of these reported not seeing a dietitian. Of 156 patients who provided feedback, almost all reported that the MST was clear (92%), questions were easy to understand (95%), and completion time was ≤5 minutes (99%).

Conclusion

Patient-led screening with the MST is reliable and well accepted by patients. Patient-led screening in the cancer care ambulatory setting has the potential to improve patient autonomy and screening completion rates.

中文翻译：

---

使用营养不良筛查工具进行患者主导的筛查的可靠性：在癌症护理非卧床环境中，患者与医疗保健专业评分之间的协议

背景

据报道，癌症患者营养不良的患病率高达60％至80％，营养不良与较低的生存率，对治疗的反应减少以及功能状态较差有关。营养不良筛查工具（MST）是由医疗保健专业人员管理的经过验证的工具；但是，尚未对其进行患者主导的筛查进行评估。

目标

本研究旨在通过评估患者主导和营养学家研究者主导的筛选之间的评估者间可靠性以及初次和重复患者筛选之间的评估者内部可靠性来评估患者主导的MST筛选的可靠性。

设计/参加者

这项横断面研究包括2016年10月在澳大利亚昆士兰州大都会教学医院接受门诊癌症护理服务的208位成年人（n = 160评价者间可靠性; n = 48评估者内部可靠性在单独的样本中测得）。

主要结果测量

主要结局指标为MST危险类别（MST 0-1：无危险，MST≥2：有危险），由患者和营养师完成的筛查，患者试验-再筛查和患者可接受性确定。

统计分析

百分比和机会校正协议（Cohen卡帕系数，κ）用于确定患者-MST与营养师-MST（评分间信度）之间的一致性，以及患者入院时所完成的MST（患者-MSTA）之间的一致性，以及由患者完成的MST之间的一致性患者在完成初始MST（患者MSTB）后1到3个小时（评分者可靠性）。

结果

观察者之间的信度和信度内部的信度都很高。患者-MST与营养师-MST之间的一致性为96％，机会调整后的一致性“几乎完美”（κ= 0.92，95％CI为0.84至0.97）。重复的患者-MSTA和患者-MSTB之间的一致性为94％，机会调整后的一致性“几乎完美”（κ= 0.88，95％CI为0.71至1.00）。根据营养师MST，确定有33％（n = 53）的患者有营养不良的风险，其中40％的人报告没有营养师。在156位提供反馈的患者中，几乎所有患者均报告说MST明确（92％），问题易于理解（95％），完成时间≤5分钟（99％）。

结论

用MST进行患者主导的筛查是可靠的，并已为患者所接受。在癌症护理门诊环境中由患者主导的筛查有可能提高患者的自主权和筛查完成率。