Departments of Epidemiology and Biostatistics (AV, MA, SC), Anesthesiology (PD, TM, BS) and Surgery (MH, JC, SC), Memorial Sloan Kettering Cancer Center, 485 Lexington Ave, 2Nd Floor, New York, NY, 10017, USA.
Department of Urology, (SC) Sahlgrenska Academy at Gothenburg University, Gothenburg, Sweden.
Health Qual Life Outcomes. 2023 Sep 13;21(1):105. doi: 10.1186/s12955-023-02184-0.
Verbal rating scales (VRS) are widely used in patient-reported outcome (PRO) measures. At our institution, patients complete an online instrument using VRSs with a five-point brief response scale to assess symptoms as part of routine follow-up after ambulatory cancer surgery. We received feedback from patients that the brief VRS descriptors such as "mild" or "somewhat" were vague. We added explicit descriptors to our VRSs, for instance, "Mild: I can generally ignore my pain" for pain severity or "Somewhat: I can do some things okay, but most of my daily activities are harder because of fatigue" for fatigue interference. We then compared responses before and after this change was made.
The symptoms investigated were pain, fatigue and nausea. Our hypothesis was that the explicit descriptors would reduce overall variance. We therefore compared the coefficient of variation of scores and tested the association between symptoms scores and known predictors thereof. We also compared time to completion between questionnaires with and without the additional descriptors.
A total of 17,500 patients undergoing 21,497 operations were assigned questionnaires in the period before the descriptors were added; allowing for a short transition period, 1,417 patients having 1436 operations were assigned questionnaires with the additional descriptors. Symptom scores were about 10% lower with the additional descriptors but the coefficient of variation was slightly higher. Moreover, the only statistically significant difference between groups for association with a known predictor favored the item without the additional language for nausea severity (p = 0.004). Total completion time was longer when the instrument included the additional descriptors, particularly the first and second time that the questionnaire was completed.
Adding descriptors to a VRS of post-operative symptoms did not improve scale properties in patients undergoing ambulatory cancer surgery. We have removed the additional descriptors from our tool. We recommend further comparative psychometric research using data from PROs collected as part of routine clinical care.
在患者报告的结果（PRO）测量中，广泛使用口头评分量表（VRS）。在我们的机构中，患者在门诊癌症手术后的常规随访中，使用 VRS 完成带有五点简短反应量表的在线工具，以评估症状。我们收到患者的反馈，称简短的 VRS 描述词，如“轻度”或“有些”，有些模糊。我们在 VRS 中添加了明确的描述词，例如“轻度：我通常可以忽略我的疼痛”用于疼痛严重程度，或“有些：我可以做一些事情，但由于疲劳，我的大部分日常活动都更困难”用于疲劳干扰。然后，我们比较了更改前后的反应。
调查的症状包括疼痛、疲劳和恶心。我们的假设是，明确的描述词会降低整体方差。因此，我们比较了分数的变异系数，并测试了症状分数与已知预测因素之间的关联。我们还比较了带有和不带有附加描述符的问卷之间的完成时间。
在添加描述符之前，共有 17500 名接受 21497 次手术的患者被分配了问卷；在允许短暂过渡的情况下，有 1417 名接受了 1436 次手术的患者被分配了带有附加描述符的问卷。添加描述符后，症状评分约降低了 10%，但变异系数略高。此外，与已知预测因素的关联中，唯一具有统计学意义的组间差异有利于没有附加语言的恶心严重程度项目（p=0.004）。当仪器包含附加描述符时，总完成时间会更长，尤其是当问卷第一次和第二次完成时。
在接受门诊癌症手术的患者中，给术后症状的 VRS 添加描述词并没有改善量表特性。我们已经从我们的工具中删除了额外的描述符。我们建议使用作为常规临床护理一部分收集的 PRO 数据进行进一步的比较心理测量学研究。
