营养风险筛查2002改善临床结局

Screening by NRS 2002 improves clinical outcome

评价一种筛查工具的预测有效性（validity），需观察经该工具筛查阳性的患者接受治疗后，能否改善临床结局。尤其在临床营养中，仅基于观察性研究所获得的工具，其筛查阳性结果不足以反映对不良结局的预测。因为在许多疾病过程中，营养不足和并发症常常相伴而行，且营养不足未必是导致这些并发症的因素。因此，开发一种可供医院和其他诊疗机构使用的、具备良好有效性的营养风险筛查工具绝非易事，因为迄今为止仅有很少的随机对照研究（RCT）是基于与结局相关的营养筛查而纳入患者的。另外，针对患者诊疗质量管理以及针对营养制剂报销制度的政策均依据营养风险筛查。过去10～20年也出台了几种筛查工具，而到底该选用何种工具存在较大困惑。笔者指出，“营养风险筛查2002”（NRS 2002）是有效性最佳的筛查工具，理由如下：（1）包括了所有被广泛认同的疾病相关营养不足的评分指标，及因营养不足未经治疗而导致并发症的风险。该评分系统既包括了营养状况评分，如体重指数（BMI）、近期体重下降和近期摄食量减少。（2）将上述指标纳入数值评分系统。（3）其有效性已在观察性结局研究、系统文献分析、随机对照研究和大规模前瞻性队列研究中获得证实。笔者就NRS 2002的预测有效性与其他筛查或评定工具进行比较，着重评价其在癌症患者和老年住院患者中的应用价值。并在新近国际共识的框架下，就疾病相关营养不足的诊断进行论证。

Predictive validity of a screening tool requires that clinical outcome improves when those screened positive are treated according to the result of the screening.Particularly in clinical nutrition, predictive validity only in the form of observation studies showing adverse outcome in those screened positive is insufficient since malnutrition and complications go hand in hand in many disease processes without malnutrition necessarily being the intermediate cause of the complication.It is therefore not a simple task to develop a well validated screening tool for nutritional risk in hospitals and other care settings, since very few randomized controlled trials are available in which patients were included on the basis of outcome by nutritional screening. On the other hand, regulations for quality management of patient care and rules for reimbursement of nutritional products require screening to be in place.The result is the development of several screening tools in the last 10-20 years and a great deal of confusion about which tool to use. In this editorial it is argued that Nutritional Risk Screening 2002 （NRS 2002） is the best validated screening tool since （1） it includes all the widely accepted components of disease related malnutrition, or rather risk of developing complications due to untreated malnutrition： a score for nutritional status including BMI, recent weight loss and recent food intake in combination with an estimate of possibly increased nutritional requirements, （2） it links these components to a numerical scoring system and （3） it has been validated in observational outcome studies, by a systematic literature analysis, in randomized controlled trials and in large scale prospective cohort studies.The predictive validity of NRS 2002 is discussed in relation to other screening or assessment tools with specific comments on its use in cancer patients and in hospitalized elderly. It is discussed within the framework of the recent international consensus on etiology and diagnosis of disease-related malnutrition.