基于Donabedian模型的脑卒中后吞咽困难护理管理能力量表的开发与心理测量学评价

文献信息

中文摘要

背景： 脑卒中后吞咽困难常见，并对患者结局产生不良影响，但目前尚无标准化工具用于评估该领域的护理管理能力。
目的： 开发并验证脑卒中后吞咽困难护理管理能力（NMC-PSD）量表，这是一种自我报告工具，用于测量护士对病区层面管理能力的感知。
方法： 该35条目量表改编自现有核查表，并通过两轮德尔菲咨询（I-CVI=0.80-1.00；S-CVI/Ave=0.93）和预试验进行完善。一项横断面调查提供了心理测量学验证。该量表采用混合模型：形成性结构维度（资源投入共同构成能力）与反映性过程维度和结果维度并列。反映性维度采用Cronbach's α评估信度，总量表及人力资源复合指标采用重测ICC评估信度。
结果： 在来自中国12个省份的410名护士中，26条目反映性核心显示出足够的内部一致性（过程α=0.97；结果α=0.95）和重测信度（ICC=0.98）。双因子模型拟合良好：χ2(348)=699.11，CFI=0.958，TLI=0.951，RMSEA=0.050，SRMR=0.027。26条目反映性核心显示出占主导的单维性（ECV=0.869，ωH=0.963），并被推荐作为主要验证指标（范围26-130）。九个形成性结构指标得到内容效度支持（S-CVI/Ave=0.93），作为描述性资源投入单独报告。六个特定因子的独立信度较低（ωHS=0.05-0.19），仅作为概念性或描述性分组。反映性总分均值为54.71（SD=18.40）；探索性加权指数为207.87（SD=66.41）。
结论： NMC-PSD量表作为护士感知病区能力的个体层面测量工具，显示出可接受的心理测量学特性。在同一研究样本中事后进行的双因子重新设定揭示了26个反映性条目背后占主导的一般因子。有必要进行独立交叉验证。推荐将26条目反映性总分作为主要验证指标；九个形成性结构指标作为描述性资源投入单独报告。分量表剖面仅作为概念性或描述性条目分组。
对护理管理的意义： 26条目反映性总分支持对护士病区管理能力感知的个体层面推断。组织层面基准比较需要进一步验证病区内一致性和聚合信度。九个结构指标提供描述性资源画像，探索性加权指数在针对客观管理指标进行效标关联验证之前仍属补充性指标。

英文摘要

BACKGROUND: Poststroke dysphagia is common and adversely affects patient outcomes, yet no standardised tool exists to assess nursing management capacity in this area.
AIM: To develop and validate the nursing management capacity in poststroke dysphagia (NMC-PSD) Scale, a self-report instrument measuring nurses' perception of unit-level management capacity.
METHODS: The 35-item scale was adapted from an existing checklist and refined through two rounds of Delphi consultation (I-CVI = 0.80-1.00; S-CVI/Ave = 0.93) and pilot testing. A cross-sectional survey provided psychometric validation. The scale uses a mixed model: a formative structure dimension (resource inputs jointly constituting capacity) alongside reflective process and outcome dimensions. Reliability was evaluated using Cronbach's α for the reflective dimensions and test-retest ICC for the total scale and the human resources composite.
RESULTS: In 410 nurses from 12 Chinese provinces, the 26-item reflective core showed adequate internal consistency (process α = 0.97; outcome α = 0.95) and test-retest reliability (ICC = 0.98). Bifactor model fit was good: χ2(348) = 699.11, CFI = 0.958, TLI = 0.951, RMSEA = 0.050, SRMR = 0.027. The 26-item reflective core showed dominant unidimensionality (ECV = 0.869, ωH = 0.963) and is recommended as the primary validated metric (range 26-130). Nine formative structural indicators, supported by content validity (S-CVI/Ave = 0.93), are reported separately as descriptive resource inputs. Six specific factors had low independent reliability (ωHS = 0.05-0.19) and serve as conceptual or descriptive groupings only. Mean reflective total score was 54.71 (SD = 18.40); the exploratory weighted index was 207.87 (SD = 66.41).
CONCLUSIONS: The NMC-PSD Scale demonstrated acceptable psychometric properties as an individual-level measure of nurse-perceived unit capacity. A bifactor re-specification, developed post hoc within the same sample, revealed a dominant general factor underlying the 26 reflective items. Independent cross-validation is warranted. The 26-item reflective total score is recommended as the primary validated metric; nine formative structural indicators are reported separately as descriptive resource inputs. Subscale profiles serve as conceptual or descriptive item groupings only.
IMPLICATIONS FOR NURSING MANAGEMENT: The 26-item reflective total score supports individual-level inference regarding nurses' perceptions of unit management capacity. Organisational-level benchmarking requires further validation of within-unit agreement and aggregation reliability. The nine structural indicators offer descriptive resource profiling, and the exploratory weighted index remains supplementary pending criterion-related validation against objective administrative indicators.