用于犬皮肤马拉色菌定量的胶带粘贴细胞学镜检的重复性...

王帆

Repeatability and reproducibility of microscopic examination of adhesive tape strip cytology slides for the quantification of Malassezia spp. in canine skin

用于犬皮肤马拉色菌定量的胶带粘贴细胞学镜检的重复性和再现性研究

翻译：林晓丹 校对：王帆

Abstract

摘要

Background: The optimal microscopic magnification and number of optical fields of adhesive tape strip cytological slides that should be examined when searching for Malassezia yeasts on canine skin are unknown.

Objectives: To determine the optimal magnification and the minimum number of optical fields that should be examined to maximise intraobserver repeatability and interobserver reproducibility.

Materials and Methods: Seven experienced examiners counted, twice, the number of yeasts in 10, 20, 30, 40 and 50 optical fields of 40 slides at ×400 and ×1000 magnification.

Results: The number of yeasts per unit surface area was significantly higher at ×1000 compared to ×400 magnification. Repeatability and reproducibility for counting the yeasts was very poor.

Conclusions and clinical relevance: Adhesive tape strip cytological slides should be examined microscopically for Malassezia spp. at ×1000 magnification. The repeatability of this examination for counting the yeasts is poor.

KEYWORDS cytology, Malassezia, repeatability, reproducibility

背景：在犬皮肤上寻找马拉色菌时，胶带粘贴细胞学镜检的最佳显微放大倍数和光场数量尚不清楚。

目标：确定最佳的放大倍数和应检查的最小光场数，以最大限度地提高观察者内的重复性和观察者间的再现性。

材料和方法：7名经验丰富的审查员在400倍和1000倍放大下，在10、20、30、40和50个光场中计算酵母菌数量。

结果：与400倍放大相比，1000倍下单位表面积的酵母菌数量显著增加。酵母菌计数的重复性和再现性都很差。

结论及临床相关性：胶带粘贴细胞学应在显微镜1000倍下检查马拉色菌。这种酵母菌计数方法的重复性较差。

关键词：细胞学、马拉色、重复性、再现性

INTRODUCTION

介绍

Yeasts of the genus Malassezia, especially M. pachydermatis, are members of the normal flora of canine skin, where they are usually present in low numbers. Increased population and/or development of hypersensitivity to their allergens commonly occur secondarily to other cutaneous or systemic diseases, and result in dermatitis that is typically manifested by pruritus, erythema, hyperpigmentation, scaling, crusting, lichenification and thickening of the skin.

马拉色菌属的酵母菌，特别是厚皮马拉色菌，是犬皮肤正常菌群的成员，它们通常数量很少。马拉色菌数量增加和/或对过敏原过敏反应通常继发于其他皮肤或全身性疾病，并导致以瘙痒、皮肤发红、色素沉着、皮屑、结痂、苔藓化和皮肤增厚为明显表现的皮肤病。

In clinical practice, a suspicion of Malassezia dermatitis, in dogs with compatible clinical signs, usually is followed by microscopic examination of cytological slides. The same examination has been proposed for monitoring antifungal treatment, and this application may become more important in the future if drugresistant strains become widespread.Furthermore, trials on the efficacy of various treatments for canine Malassezia dermatitis have been based on the quantitative or semiquantitative microscopic examination of cytological slides. Adhesive tape strip is the most commonly used sampling technique. However, the optimal microscopic magnification, that may be either ×400 (high-power; HP) or ×1000 (oil immersion; OI),and the number of optical fields that should be examined have not been standardised, even though they may have a profound impact on the sensitivity, specificity, repeatability and reproducibility of the examination.

在临床实践中，怀疑马拉色菌皮肤病的犬，通常是显微镜检查细胞压片检查。同样的检测也被提出用于监测抗真菌治疗，如果耐药菌株的种类更加广泛，这种应用在未来可能变得更加重要。此外，犬马拉色菌皮肤病各种治疗疗效的试验是基于细胞学压片的定量或半定量显微镜检查。胶带粘贴是最常用的取样技术。然而，最佳的显微镜放大倍数，可能是×400(高功率；HP)或×1000(油浸；OI)，和应该检查的光场的数量都还没有被标准化， 尽管它们可能对检查的敏感性、特异性、重复性和再现性有深远的影响。

The aim of this study was to find the optimal microscopic magnification and the minimum number of optical fields that should be examined to achieve the best possible intraobserver repeatability and interobserver reproducibility for the quantification of Malassezia spp. yeasts in adhesive tape strip cytological slides obtained from the skin of dogs suspect of having Malassezia dermatitis.

本研究的目的是找到最佳的显微镜放大倍数和最小的光场数，以达到最佳的观察者内重复性和观察者间再现性，用于疑似患有马拉色菌皮肤病的犬的皮肤上获得的胶带粘贴细胞学检查中的马拉色酵母菌定量。

MATERIALS AND METHODS

材料和方法

Examiners, adhesive tape strip cytological slides and microscope

审查员、胶带粘贴法细胞学检查和显微镜

A total of seven examiners participated in the study. They included one European College of Veterinary Dermatology board-certified first opinion veterinary dermatological practitioner (VD; Examiner 1) and six veterinarians with many years of experience in cutaneous cytology, including clinic staff members, PhD students and a dermatology intern (examiners 2–7).

共有7名审查员参与了这项研究。他们之中包括一个欧洲兽医皮肤科学院委员会认证的第一意向兽医皮肤科医生（VD；审查员1）以及6名具有多年皮肤细胞学经验的兽医，包括诊所工作人员、博士生和一名皮肤科实习生（审查员2-7）。

The adhesive tape strip cytological slides that were used had been collected by the VD during the previous year and stored for the purposes of this study. They had been obtained during routine clinical work-up, using approximately 5cm long pieces of clear adhesive tape (Scotch Crystal Clear Tape, 3 M; Woodmead, South Africa), pressed 5–10 times onto the lesional skin of dogs suspected of Malassezia dermatitis, presenting pruritus and/or compatible skin lesions.The tape was stained with Diff-Quik1 (Merck; Darmstadt, Germany), using the fixative and both staining solutions, air-dried and pressed, with the sticky side down, onto a clean microscope slide. Before the beginning of the study, the VD checked all slides to ensure that they retained acceptable staining quality (i.e. intense basophilic staining of nuclear material) and none of them had to be discarded

所使用的胶带粘贴细胞学切片已由VD在前一年收集并存储，用于本研究的目的。它们是在常规临床检查中获得的，使用大约5厘米长的透明胶带，在表现为瘙痒和/或并发皮肤病变，疑似马拉色菌皮肤病的犬的皮肤病变上按压5-10次。胶带用Diff-Quik染色，使用固定剂和两种染色溶液，风干，粘面向下压在一个干净的显微镜载玻片上。 在研究开始前，VD检查了所有的玻片，以确保它们保持可接受的染色质量（即有核物质的强嗜碱性染色），没有一个是必需被丢弃的。

All microscopic examinations were performed using the same microscope (ICS-KF2 Binocular Research Microscope, Carl Zeiss; Jena, Germany) with two ×10/18 eyepieces.

所有显微镜检查均使用同一显微镜，两个目镜10倍放大/18。

Training of the examiners

审查员培训

One week before the beginning of the study, the examiners attended a 60min training course that was run by the VD. Initially, a digital slide presentation was used to explain the aims and the design of the study, how to use the microscope, how to select optical fields (choosing optical fields with a moderate to high numbers of keratinocytes in monolayer located in the central area of the slide was advised because, during sampling, the centre of the adhesive tape was opposed to the skin lesion), how to identify Malassezia spp., and which are the common clinical manifestations of canine Malassezia dermatitis. This was followed a live demonstration of microscopic identification and counting of Malassezia spp., using a BX-40 microscope equipped with Altra 20 screenprojecting colour camera (Olympus; Tokyo, Japan).

在研究开始前一周，审查员参加了由VD举办的60 分钟培训课程。最初，我们使用数字幻灯片演示来解释研究的目的和设计，如何使用显微镜，如何选择光场(建议选择位于载玻片中心区域的单层至中等到大量角质形成细胞的光场，因为在取样过程中，胶带的中心与皮肤病变区相对应)，如何鉴别马拉色菌，还有犬马拉色菌皮肤病的常见临床表现。随后是对马拉色菌的显微镜鉴定和计数的现场演示，使用配备有Altra20屏幕投影彩色相机的BX-40显微镜。

Microscopic examination

显微镜检查

From a collection of 100 adhesive tape strip cytological slides, 40 were selected randomly, and with their identity masked, they were numbered randomly from 1 to 40. Each examiner was randomised to examine these 40 slides either first at HP, then at OI, then again at HP and finally at OI, or in the reverse order. For each examiner, a period of at least 1week lapsed between the end of each examination round and the beginning of the next round. During the examination of each slide,the number of Malassezia spp. counted in the first 10, 20, 30, 40 and 50 optical fields and the time spent was recorded.

从100张胶带粘贴细胞学切片中，随机选择40张，并掩盖其标注，随机编号从1到40。每个审查员被随机地检查这40张载玻片，首先在高倍镜，然后在油镜，然后在高倍镜，最后在油镜，或者以相反的顺序，对于每个审查员来说，从每一轮考试结束到下一轮考试开始之间至少有一周的时间。在检查每张载玻片中马拉色菌的数量时，在前10、20、30、40和50个光场中进行计数，并记录所花费的时间。

Statistical methods

统计方法

For all randomisations, a freely available random number generator software(https://www.gigacalculator.com/c ... umber-generator.php) was used, statistical analyses were performed in Spss 21 for Windows and the level of significance was set at 5%.

对于所有的随机化方法，都有一个免费提供的随机数生成器软件，在Windows Spss21中进行统计分析，显著性水平设置为5%。

The distribution of all continuous variables was tested using the Lilliefors modification of the Kolmogorov– Smirnov test.

所有连续变量的分布均采用Kolmogorov– Smirnov修正的Lilliefors检验进行检验。

In order to compare the number of Malassezia spp. between the two microscopic magnifications, the number of yeasts at OI magnification was multiplied by 6.25 to correct for the smaller microscopic optical field surface area and this product was compared with the number of yeasts at HP magnification, separately for each examiner, for each of the number of optical fields examined (10, 20, 30, 40 or 50) and for each examination row (first or second). For these 70 comparisons, the distribution of the number of yeasts did not follow a normal distribution (p<0.001), and thus the related sample Wilcoxon signed-rank test was used.

为了比较两种显微放大倍数下马拉色菌的数量，在油镜放大倍数下的酵母菌数乘以6.25，以校正较小的显微镜光场表面积，并将该结果与高倍镜放大倍数下的酵母菌数量进行了比较，分别为每个审查员，为每个检查的光场数（10、20、30、30、40或50）以及每个检查行（第一或第二）。在这70个比较中，酵母菌数量的分布并不遵循正态分布(p<0.001)，因此，相关样本采用Wilcoxon符号秩检验。

For the evaluation of intraobserver repeatability, the difference between the two measurements of the number of Malassezia spp. for the same cytological slide, the same examiner, the same number of optical fields and the same magnification was calculated. None of the 70 datasets of differences followed the normal distribution (all p-values<0.001). Subsequently, the null hypothesis that the median of the differences (d) was zero, was tested using the Wilcoxon signed-rank test and could not be rejected (p≥0.071) for 66 of 70 datasets. The standard deviation of the differences (SDdif) and the intraclass correlation coefficients (ICC) of the datasets were calculated and compared not only among the different number of optical fields at the same microscopic magnification, but also between the two microscopic magnifications for the same number of optical fields, using either a paired sample Student's t-test (normal distribution) or related sample Wilcoxon signed-rank test (non-normal distribution). The results were evaluated after Bonferroni correction for multiple comparisons. Finally, the differences of the two measurements were plotted against their mean, separately for each number of optical fields×microscopic magnification combination.

为了评估观察者内的重复性，对于相同的细胞学切片，相同的审查员，相同数量的光场和相同的放大倍数下，两种马拉色菌数量之间的差异。70个差异数据集中没有一个遵循正态分布(均为p值<0.001)。随后，采用Wilcoxon符号秩检验，假设差异的中位数（d）为零，在70个数据集中，有66个不能采纳(p≥0.071)。使用配对样本Student’st检验(正态分布)或相关样本Wilcoxon符号秩检验(非正态分布)，计算和比较数据集的差异标准差(SDdif)和类内相关系数(ICC)，不仅在相同显微镜放大倍数下的不同光场数之间，而且在相同光场数下的两种显微镜放大倍数之间。在Bonferroni校正多次比较后，对结果进行评估。最后，将两种测量结果的差异与它们的平均值绘制成图，分别为不同光场数×显微放大倍数组合。

For the evaluation of the interobserver reproducibility, the difference between the measurements of the number of Malassezia spp. by each pair of examiners for the same cytological slide, at the same round, the same number of optical fields and the same magnification was calculated.

为了评估观察者间的再现性，每一对审查员对相同的细胞学切片，在同一轮中，计算相同数量的光场和相同的放大倍数下，马拉色菌数量之间的数值差异。

The time to examine the cytological slides and count Malassezia spp. was pooled for both examination rounds of all examiners. Because all datasets did not follow the normal distribution (all p-values<0.001) the median value of examination time was compared between the two microscopic magnifications for 10, 20, 30, 40 or 50 optical fields using the Wilcoxon matchedpair signed-rank test.

所有审查员的细胞学切片检查和马拉色菌计数时间合并。由于所有数据集都不遵循正态分布(所有p值<0.001)，使用Wilcoxon 符号秩检验对比了10个、20个、30个、40个或50个光场，在两种显微镜放大倍数下检查时间的中位数。

Ethics

伦理标准

Owing to the retrospective use of adhesive tape strip cytological slides that had been obtained for routine diagnostic purposes, no ethical approval was necessary.Moreover, sampling was noninvasive, assuring limited animal distress.

由于回顾性使用了用于常规诊断目的的胶带粘贴细胞学检验，因此不需要伦理批准。此外，取样是无创的，确保有限的动物痛苦。

RESULTS

结果

The median value of the total number of Malassezia spp. corrected for the microscopic optical field surface area was significantly higher for OI compared to HP magnification, except for one examiner (Examiner 3) (see Table S1). For this examiner, in the first round of microscopic examinations there was no significant difference in the total number of Malassezia spp. irrespective of the number of optical fields examined (10, 20, 30, 40 or 50). In the second round there was no difference at 10 optical fields, the number of Malassezia spp. was higher for HP at 20 optical fields, and higher for OI at 30, 40 and 50 optical fields. Also, a systematic difference was found for the same examiner when counting Malassezia spp. in 10 (p = 0.033), 30 (p = 0.014), 40 (p = 0.008) and 50 (p = 0.004) HP fields. Therefore, Examiner 3 was considered an outlier and excluded from further analysis.

经显微镜光场表面积矫正后额马拉色菌总数的中位数，除了一名审查员外（审查员3），油镜数显著高于HP（见表S1）。在第一轮显微镜检查中，不管检测的光场数如何（10、20、30、40或50）马拉色菌的总数无显著性差异。在第二轮试验中，在10个光场上，马拉色菌的数量没有差异。在20光场时高倍镜更高，在30、40和50光场时油镜更高。此外，在计算马拉色菌时也发现了系统性差异。在高倍镜视野下10 (p = 0.033), 30 (p = 0.014), 40 (p = 0.008) and 50 (p = 0.004)，因此，审查员3被认为是一个离群值，并被排除在进一步的分析之外。

Intraobserver repeatability

观察者内重复性

the SDdif and ICC of the differences between the two measurements of the number of Malassezia spp. for the same cytological slide, the same examiner, the same number of optical fields and the same magnification are presented on Table S2. When the results of the six examiners were pooled, the comparisons of SDdif and ICC among 10, 20, 30, 40 and 50 optical fields at the same microscopic magnification (×400 or ×1000),there were no significant differences after Bonferroni correction for multiple comparisons (Figure 1; Table S3). Also, there were no differences in the comparisons between ×400 and ×1000 magnification for the same number of optical fields (Table S4). The plots of the differences of the two measurements against their mean showed a cone effect indicative of heteroscedasticity (Figure S1 that became more obvious when extreme values (means>100) were omitted (Figure 2). Thus, intraobserver repeatability was poor, independently of the number of optical fields and of the microscopic magnification, and it became poorer at higher yeast counts.

表S2列出相同的细胞学切片，相同的审查员，相同数量的光场和相同的放大倍数下，两次测量马拉色菌数量的SDdif和ICC差异。当6名审查员的结果汇总时，在相同的显微镜放大下（400倍或1000倍）的10、20、30、40和50个光场SDdif和ICC的比较，经过Bonferroni多重比较，无显著性差异（图1；表S3）此外，在相同数量的光场下，400倍和1000倍的放大倍数之间的比较也没有差异(表S4)。两种测量值与平均值的差异图显示了锥效应表明异方差(图S1，当省略极端值（即>100）时，情况变得更加明显（图2）。因此，观察者内的重复性较差，与光场数和显微镜放大倍数无关，并且当酵母菌计数越高时，它就会变得越差。

Interobserver reproducibility

观察者间再现性

The Lilliefors modification of the Kolmogorov– Smirnov test showed that the distribution of 300 of 300 (100%) datasets of differences between the measurements of the number of Malassezia spp. by each pair of examiners for the same cytological slide, the same round, the same number of optical fields and the same magnification, did not follow normal distribution (all p-values<0.001), denoting a systematic bias (i.e. for each pair, one examiner was systemically counting more Malassezia spp. than the other). Subsequently, interobserver reproducibility was not analysed further.

对Lilliefors修正的Kolmogorov - Smirnov检验表明，300个中有300个（100%）马拉色菌数量测量值的差异。由每一对审查员检查相同的细胞学切片，同样的圆，同样数量的光场和同样的放大倍数，都没有遵循正态分布(所有p-values<0.001)，表示系统偏差(即对于每一对，一个审查员系统地计算更多的马拉色菌。比其他的要多)。随后，没有进一步分析观察者间的可重复性。

Examination time

检查时间

The median time needed to count Malassezia spp. varied from 55s (10 OI fields) up to 296.5 s (50 HP fields) and it was significantly shorted for ×1000 compared to ×400 magnification, for each number of optical fields examined (Table S5).

计算马拉色菌所需的中位数时间。从55s(10个油镜视野)到296.5s(50个高倍镜视野)，与400倍放大相比，对于每个数量的光场1000倍显著缩短(表S5)。

DISCUSSION

讨论

There is no consensus on the optimal microscopic magnification when examining cytological slides for the presence and count of Malassezia yeasts. Although current World Association for Veterinary Dermatology guidelines state that it may be preferable to examine the slides at OI,and this magnification has been used more often in (semi-)quantitative cytological studies,there are some publications where HP has been employed. Malassezia cells do not always have the easily recognisable peanut shape that is associated with their monopolar budding, yet they also can appear as round or oval organisms.In this case they may be less easily recognised and/or differentiated from extraneous materials, such as pollens or stain precipitates, especially at a lower microscopic magnification and for inexperienced examiners. Furthermore, many yeast cells do not take up enough stain to be easily recognisable at HP magnification. The above may explain why the number of yeasts per cytological slide surface area was almost always significantly higher at the OI magnification. Subsequently, this magnification should be considered optimal, despite the obvious drawback of scrutinising smaller areas of the cytological slide after examining the same number of optical fields.

在检查细胞学切片中马拉色菌的存在和计数时，对于最佳的显微镜放大倍数还没有达成共识，但目前的世界兽医皮肤病学协会的指南规定，最好是在油镜视野下检查，这种放大倍数在（半）定量细胞学研究中更常用，有部分出版物也使用高倍镜。马拉色菌细胞并不总是具有与它们的单极出芽相关的容易识别的花生形状，它们也可以作为圆形或椭圆形的生物体出现，在这种情况下，它们可能不太容易识别和/或与外来物质区分，如花粉或染色沉淀，特别是在较低的显微镜放大倍数和没有经验的审查员。此外，许多酵母细胞没有足够着染，在高倍镜放大下不容易识别，以上可以解释为什么在油镜放大倍数下，每个细胞学切片表面积的酵母数几乎总是显著较高，随后，尽管在检查相同数量的光场后，仔细检查细胞学切片的较小区域明显的缺陷。这种放大倍数应该被认为是最佳的。

The number of the optical fields that are examined is expected to have a major impact on the results of adhesive tape strip cytology. On the one hand, the greater the number of optical fields, the higher the chance to find the yeasts, especially when their number is low. Also, yeasts are not uniformly distributed over the slide and, thus, quantitative results are expected to be more accurate after examination of as many as possible optical fields. On the other hand, the examination of many optical fields is time-consuming. Therefore, a compromise is needed to obtain accurate results after a reasonable examination time. In previous studies the number of OI fields examined was five, 10,15,20, 30 or 100. The results of this study show clearly that there is tremendous interobserver variability in the estimation of Malassezia numbers, even when 50 random OI fields are examined. This variability is reduced when cytological slides are examined by the same investigator, yet the intraobserver repeatability still was found to be very low to be considered acceptable, probably because different optical fields are examined at each examination run.Our finding may explain why no significant correlation of the number of Malassezia spp. per surface area of the skin was found between quantitative culture (detergent scrub technique) and adhesive tape strip cytology (after examination of 20 OI). Also, our results are in line with a previous study, where intraobserver repeatability and interobserver reproducibility were poor (6% overall reproducibility of yeast counts) when yeast numbers were measured quantitatively in 10 OI fields of impression smears that had been obtained from skin lesions of 20 dogs with superficial pyoderma and from the dorsum of five healthy dogs. Perhaps repeatability and reproducibility could be improved if >50 optical fields are examined, but in this case the test would become more time-consuming and less practical. For these reasons, microscopic examination of adhesive tape strip cytological slides should not be used for the quantification of Malassezia spp., especially when accuracy is highly important, as in therapeutic trials. Instead, quantitative culture (e.g. detergent scrub sampling followed by inoculation onto modified Dixon's agar), or accurate quantitative molecular tests should be employed for this purpose.

所检查的光场的数量预计将对胶带粘贴细胞学的结果有很大影响。一方面，光场的数量越多，找到酵母菌的机会就越高，特别是当它们的数量较低的时候。此外，酵母菌并不均匀地分布在载玻片上，因此，在检查尽可能多的光场后，定量结果有望更准确。另一方面，对许多光场的检查是耗时的。因此，在合理的检查时间后，需要妥协准确的结果，在以往的研究中，检测的油镜视野的数量为5、10、15、20、30或者100。这项研究的结果清楚地表明，即使当50个随机油镜视野被检查时，在马拉色菌数量的估计中也存在巨大的观察者间差异。当由同一研究者检查细胞学切片时，这种可变性就减少了，然而，观察者内的可重复性仍然很低，因此被认为是可以接受的，这可能是因为在每次检查时都检查了不同的光场。我们的发现可能可以解释为什么没有显着的相关性的数字，每一皮肤表面积马拉色菌的比例为在定量培养（洗涤剂擦洗技术）和胶带细胞学(20个OI检查后)。此外，我们的研究结果与之前的研究结果一致， 从20只患有浅表性脓皮病的犬的皮肤病变和5只健康犬的背部中获得的细胞学压片，当在10个油镜视野中定量测量酵母菌的数量时，观察者内的重复性和观察者间的再现性较差（酵母菌计数的总重复性为6%）。如果检查>50光场，也许重复性和再现性可以得到提高，但在这种情况下，测试将变得更耗时和更不实用。由于这些原因，胶带粘贴细胞学压片的显微镜检查不应用于马拉色菌的定量，特别是当准确性非常重要时，比如在治疗试验中。相反，应该采用定量培养(例如，洗涤剂擦洗取样，然后接种到改良的Dixon琼脂上)、或为此目的应采用准确的分子定量检测。

A previous study showed high intraobserver repeatability (86.2±5% agreement) and interobserver reproducibility (86.4±12.5%) when the presence and numbers of yeasts were evaluated by 60 examiners in 10 cytological slides.In that study, samples were obtained not only from dogs, but also from cats, and not only from skin, but also from ear canals; skin samples were obtained with the impression smear and not the adhesive tape strip technique; the microscopic magnification and the number of optical fields that were examined were not standardised; the number of Malassezia spp. was expressed on a semiquantitative scale (absent,occasional, low, large, massive amounts); two examinations of the same slide were considered to be in agreement if they differed by one grade (except if the grade was “absent”); and few smears had yeasts (the median value was zero) which is expected to artificially inflate the calculated repeatability and reproducibility. However, in another, more recent, study the intraobserver repeatability was excellent after microscopic examination of five OI fields, when using the same semiquantitative scale to estimate yeast numbers in adhesive tape strip cytological slides obtained from the skin of dogs and cats. Subsequently, and taking into consideration our results, this semiquantitative scale should be considered preferable over counting yeast numbers.

先前的一项研究显示，当60名审查员在10张细胞学切片中评估酵母菌的存在和数量时，观察者内重复性（86.2±5%的一致性）和观察者间再现性（86.4±12.5%）都较高，在这项研究中，样本不仅来自犬，也来自猫，不仅来自皮肤，还来自耳道；皮肤样本是通过皮肤压片而不是胶带粘贴技术获得的；显微镜下的放大倍数和被检查的光场的数量没有标准化；马拉色菌的数量，在半定量评分表上表示(没有、偶见、少量、中量、大量)；如果对同一切片的两次检查分数不同，则认为是一致的（除非分数是“没有”），很少有涂片有酵母（中位数为零），预计这将人为地夸大计算的重复性和再现性。然而，在另一项最近的研究中，当使用相同的半定量评分表来估计从犬和猫的皮肤中获得的胶带粘贴细胞学压片中的酵母菌数量时，在对5个油镜视野的显微镜检查后，观察者内的重复性非常好。随后，考虑到我们的结果，半定量评分表应该被认为比酵母菌计数更可取。

There are four shortcomings of the present study:

本研究有四个不足：

1.There is a lack of a gold standard to estimate the presence and number of yeasts (i.e. detergent scrub sampling followed by inoculation onto modified Dixon's agar)1,21 that would allow the evaluation of the accuracy of microscopic quantification. However, with such a tremendous intraobserver and interobserver variability, a poor accuracy, sensitivity and specificity is to be expected for most examiners, examination rounds and number of optical fields.

1.目前缺乏一个金标准来估计酵母菌的存在和数量（即洗涤剂擦洗取样，然后接种到改良的迪克森琼脂上），这将允许显微镜定量评估的准确性，然而，有如此巨大的观察者内和观察者间的多变性，对于大多数审查员、检验和光场数来说，准确性、灵敏度和特异性较差。

2.The examination did not continue beyond 50 OI fields which could improve intraobserver repeatability and interobserver reproducibility. However, as already mentioned, examination of more OI fields would render the test time-consuming and impractical.

2.检查没有继续超过50个油镜视野，这可以提高观察者内的重复性和观察者间的再现性，然而，如前所述，检查更多的油镜视野将使测试更加耗时且不切实际。

3.The lack of examination of cytological slides obtained by different techniques, such as impression, scrape or swab smears that may have led to different results; and

3.缺乏对通过不同技术获得的细胞学压片的检查，如皮肤压片、皮肤刮片或拭子采样，这可能导致不同的结果；

4.The possible impact of differences among the examiners, in terms of their experience, visual acuity, motivation, and perhaps other factors, on the results. This may explain the lack of reliability of the results from Examiner 3, who had to be excluded from analysis mainly because of the systematic difference in Malassezia spp. counts between the first and second microscopic examination of 10, 30, 40 and 50 HP fields. However, visual inspection of the data (SDdif, ICC, plots) did not show an obvious difference between the VD and the remaining examiners, further supporting our hypothesis that the low intraobserver repeatability and interobserver reproducibility is an inherent characteristic of the microscopic examination of adhesive tape strip cytological slides for the quantification of Malassezia spp. in canine skin. It is most likely that this occurs because different optical fields are examined each time coupled with the nonuniform distribution of the yeasts on the slides.

4.审查员之间的经验、视觉敏度、动机以及其他因素方面的差异对结果的可能影响，这可能解释了来自审查员3的结果缺乏可靠性，主要因为在第一次和第二次显微镜检查10、30、40和50高倍镜视野下，马拉色菌的计数差异而被排除在分析之外。然而，对数据的目视检查并没有显示VD和其他审查员之间的明显差异，进一步支持了我们的假设，即较低的观察者内重复性和观察者间再现性是用于在犬的皮肤上马拉色菌定量的胶带粘贴细胞学显微镜检查的固有特征，这很可能是因为每次检查不同的光场，再加上酵母菌在载玻片上的不均匀分布。

CONCLUSIONS

结论

The examination of adhesive tape strip cytological slides for presence of Malassezia spp. should be conducted at ×1000 magnification. Whilst this method is convenient for a rapid patient-side, semiquantitative assessment of cutaneous Malassezia spp. populations, the poor intraobserver repeatability and interobserver reproducibility indicates that this technique should not be used when accurate counts are required.

检查胶带粘贴细胞学压片是否存在马拉色菌，应该在1000倍放大下进行。虽然皮肤马拉色菌半定量评估法方便快捷，但是由于观察者内重复性和观察者间再现性较差，说明在需要准确计数时，不应使用该技术。

         

曾医生