Associations Between Otic Cytological Findings and Clinical Severity by Pathogen Group in Canine Otitis Externa: A Retrospective Study of 318 Cases Authors: Eun‐Seo Shin, Soh‐Yoon Kang, Min‐Sun Kim, Cheol‐Yong Hwang Published: 2026 ABSTRACT ABSTRACTBackgroundOtic cytological assessment is routinely used in canine otitis externa (OE) to assess microbial burden and inflammation. Evidence that cytological findings are associated with clinical severity, particularly by pathogen group, remains limited.ObjectivesTo evaluate associations between otic cytological findings and clinical severity in canine OE, overall and by pathogen group.AnimalsThree hundred and eighteen client‐owned dogs with OE presented to a referral dermatological service.Materials and MethodsClinical severity was scored using the Otitis Index Score 3 (OTIS3; 0–12). Cytological findings were evaluated using a semiquantitative 0–3 scale for cocci, rods and Malassezia, based on predefined institutional criteria. A total cytological score (0–9) was calculated as the sum of the scores for the three groups. Neutrophil presence was also recorded.ResultsOTIS3 scores showed a moderate correlation with total cytological score (p < 0.0001; Spearman's r = 0.45). Scores increased significantly with higher cocci and rod cytological scores (Kruskal–Wallis, p < 0.0001), while Malassezia scores were not. OTIS3 scores also were significantly higher when neutrophils were present.Conclusions and Clinical RelevanceAssociations between otic cytological findings and clinical severity of OE varied by pathogen group. Greater clinical severity was associated with neutrophil presence and bacterial burden. These findings support the interpretation of cytological findings combined with OTIS3 for assessment and monitoring of canine OE.Keywords: bacteria, dogs, ear canal, inflammation, Malassezia, neutrophils Introduction Otitis externa (OE) is a common multifactorial disease in dogs with variable clinical severity [1, 2, 3, 4, 5]. The pathogenesis of OE has been described in terms of primary causes, secondary causes, predisposing factors and perpetuating factors. Primary causes, such as allergic dermatitis, may initiate ear canal inflammation which is often exacerbated by secondary microbial infections, while chronic structural changes act as perpetuating factors that complicate resolution [3, 4]. The diagnosis of OE is established through clinical history, otoscopic examination of the ear canal and cytological examination of otic discharge [6, 7]. Otic cytological assessment is routinely used in clinical practice to assess microbial burden and inflammation [8]. Cocci, rods and yeasts are the major morphological groups observed in OE, and can be differentiated by cytological characteristics [9, 10]. Inflammatory cells, primarily neutrophils, also are captured on cytological evaluation to document the inflammatory responses in canine OE [6, 11]. Within current concepts of OE, such cytological findings may provide insight into inflammatory patterns described in erythemato‐ceruminous and suppurative forms of otitis. Erythemato‐ceruminous otitis is typically characterised by ceruminous discharge with limited inflammatory cell infiltration, while suppurative otitis is characterised by purulent exudate and neutrophilic inflammation [9, 12]. These clinical and cytological patterns are increasingly considered within the framework of dysbiosis in canine OE, where microbial burden does not always correspond directly with clinical severity [9, 13]. To accurately capture these varying clinical and microbial patterns, various approaches have been proposed to objectify and standardise cytological evaluation in canine OE [8, 14, 15, 16, 17]. Semiquantitative methods grade microbial burden and record the presence of inflammatory cells and have been applied in both clinical practice and research [18, 19]. Several clinical scoring systems have been developed to assess the severity of OE. The most widely used scoring system is the Otitis Index Score 3 (OTIS3), which grades four clinical signs (erythema, oedema/swelling, erosion/ulceration and exudate) on a 0–3 scale to yield a total score of 0–12 [20]. This scoring system provides consistent grading of clinical severity and is useful for evaluating treatment response [13, 18, 19, 21]. Although otic cytological findings and clinical severity scores are widely used in the evaluation of OE, direct comparisons between the two indicators remain limited. Particularly, pathogen‐specific associations have not been systematically investigated, and the contribution of neutrophils to disease severity is not clearly defined. The aim of this retrospective study was to investigate associations between cytological findings and clinical severity in canine OE, overall and by pathogen group, and to assess the relationship between neutrophil presence and disease severity. Materials and Methods Ethics Owner consent for the use of medical records was obtained at the time of hospital admission. Study Design and Case Selection This retrospective study included 318 client‐owned dogs that presented with OE to the dermatology service at Seoul National University Veterinary Medical Teaching Hospital (SNU‐VMTH), between January 2020 and December 2024. The diagnosis of OE was confirmed based on the patient's history, clinical signs, otoscopic examination and otic cytological findings. Dogs were eligible for inclusion if complete OTIS3 and cytological data were available at the time of the initial visit. Cases were excluded if cytological samples were missing or if only follow‐up visits were recorded without baseline data. In unilateral cases, the affected ear was included in the analysis. In bilateral cases, the ear with higher severity (higher OTIS3 or stronger cytological confirmation) was included. Clinical Severity Scoring Otoscopic examination was performed using a video‐otoscope to allow full visualisation of the vertical and horizontal ear canals [13]. The video‐otoscope was used without sedation in awake dogs. The severity of otitis was then evaluated using the OTIS3 scoring system before ear cleaning or treatment [20]. Scoring was performed at the first dermatology visit by multiple veterinary surgeons specialised in dermatology, under the supervision of a board‐certified referral clinician (DipAiCVD). All evaluators had received standardised training in the OTIS3 scoring system to ensure consistency in scoring. Cytological Assessment Ear cytological samples were collected from the junction of the vertical and horizontal ear canals using a sterile cotton‐tipped swab [7]. Samples were gently rolled onto glass microscope slides and air‐dried. Smears were then heat‐fixed before staining [8, 22]. Slides were stained using a modified Wright's stain (Diff‐Quik; Baxter Diagnostics AG). Each slide was scanned at low magnification (×4 and ×10 objectives) to identify regions with the greatest microbial or inflammatory burden. Ten nonoverlapping oil‐immersion fields (×100 objectives; ×1000 total magnification) were then selected from these regions for quantification. Counts per field for cocci, rods and Malassezia were averaged to derive a representative value, following previously described procedures [8, 14]. Average counts were assigned a semiquantitative score (0–3) according to predefined institutional criteria (Table 1). A total cytological score (0–9) was calculated as the sum of the three category scores. Because multiple microbial types could be present in the same ear, ears were included in the analyses for each corresponding cytological category. The presence or absence of neutrophils also was recorded. Neutrophils were recorded only when cells with identifiable nuclear morphology were present, and extracellular nuclear material or DNA debris was excluded. Cytological evaluation was performed independently of OTIS3 scoring during routine clinical assessment. TABLE 1Cytological scoring criteria for cocci, rods and Malassezia used in this study.ScoreCocci, rods (per OIF) Malassezia (per OIF)00–1 organismNone observed12–10 organisms1–2 yeasts211–40 organisms3–8 yeasts3≥ 41 organisms≥ 9 yeasts Note: Scores were assigned based on the number of cocci, rods and Malassezia observed per oil immersion field (OIF; ×1000). Detailed definitions are summarised in the table. Cytological scoring criteria for cocci, rods and Malassezia used in this study. Note: Scores were assigned based on the number of cocci, rods and Malassezia observed per oil immersion field (OIF; ×1000). Detailed definitions are summarised in the table. Statistical Analysis Statistical analyses were performed using prism (v10.5.0; GraphPad Software). Spearman's rank correlation was used to assess the association between OTIS3 scores and total cytological score, defined as the sum of cocci, rods and Malassezia scores. OTIS3 scores were compared across cytological grades (0–3) within each pathogen group using the Kruskal–Wallis test, followed by Dunn's—test. For comparisons between specimens with and without neutrophils at each cytological grade, Wilcoxon–Mann–Whitney U tests were performed with false discovery rate (FDR) correction to control for multiple comparisons. Statistical significance was set at p < 0.05 for Spearman's rank correlation, Kruskal–Wallis test and Dunn's—test, and q < 0.05 for FDR‐adjusted Wilcoxon–Mann–Whitney U tests. Results Study Population A total of 318 client‐owned dogs diagnosed with OE were included in the study. The median age was 8 years (range 1–16 years), and the mean age was 8.0 ± 3.4 years. The sex distribution was balanced. The most common breeds were Maltese (20.8%), poodle (16.0%), shih tzu (10.1%) and cocker spaniel (9.7%). Detailed information on sex and breed distribution is provided in Tables S1 and S2 in the Supporting Information. The overall distribution of clinical and cytological scores across the study population is summarised in Tables 2 and 3. The median OTIS3 score was 2 (interquartile range [IQR], 1–3), and the median total cytological score was 4 (IQR, 3–6). TABLE 2Frequency distribution of Otitis Index Score 3 (OTIS3) in 318 dogs diagnosed with otitis externa.OTIS3 scoreNumber of dogsProportion (%)061.918827.7210031.434714.843410.75268.2682.5761.9820.6910.310–1200.0Total 318 100 Note: Scores were assigned based on clinical evaluation of the external ear canal using the OTIS3 criteria (0–12). Frequency distribution of Otitis Index Score 3 (OTIS3) in 318 dogs diagnosed with otitis externa. Note: Scores were assigned based on clinical evaluation of the external ear canal using the OTIS3 criteria (0–12). TABLE 3Frequency distribution of total cytological score in 318 dogs diagnosed with otitis externa.Total cytological scoreNumber of dogsProportion (%)0113.51288.823210.138025.244614.55309.465918.67206.3892.8930.9Total 318 100 Note: Scores were assigned based on the sum of the category scores (0–3 each) for cocci, rods and Malassezia. Frequency distribution of total cytological score in 318 dogs diagnosed with otitis externa. Note: Scores were assigned based on the sum of the category scores (0–3 each) for cocci, rods and Malassezia. Correlation Between OTIS3 and Total Cytological Score A positive correlation was observed between OTIS3 score and total cytological score, defined as the sum of cocci, rods and Malassezia scores (p < 0.0001; Spearman's r = 0.45) (Figure 1). This indicates a moderate association between clinical severity and cytological burden. Scatter plot showing the correlation between Otitis Index Score 3 (OTIS3) and total cytological score (sum of cocci, rods and Malassezia) in 318 dogs with otitis externa. A moderate positive correlation was observed (p < 0.0001; Spearman's r = 0.45). Each dot represents one or more dogs sharing identical combinations of OTIS3 and total cytological scores. OTIS3 Across Cytological Scores by Pathogen Group OTIS3 scores differed significantly across cytological scores for both cocci and rods (Kruskal–Wallis test, p < 0.0001 for each; Figure 2a,b). For cocci, Dunn's post hoc test revealed significant pairwise differences between 0 versus 2 (p = 0.0036), 0 versus 3 (p < 0.0001) and 1 versus 3 (p = 0.0010), while the remaining comparisons were not significant. For rods, significant pairwise differences were identified between 0 versus 2 (p = 0.0298), 0 versus 3 (p < 0.0001) and 1 versus 3 (p = 0.0001), with no significant differences in the other comparisons. Otitis Index Score 3 (OTIS3) according to cytological scores (0–3) for (a) cocci, (b) rods and (c) Malassezia in 318 dogs with otitis externa. Significant differences across cytological scores were identified for cocci and rods (Kruskal–Wallis test, p < 0.0001 for both), followed by Dunn's post hoc test. No significant differences were observed for Malassezia (p = 0.0818). Significance levels: *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001; comparisons without an asterisk are not significant. By contrast, OTIS3 scores did not differ significantly across Malassezia cytological scores (Kruskal–Wallis test, p = 0.0818; Figure 2c). For Malassezia, Dunn's post hoc test revealed no significant pairwise differences (all adjusted p > 0.05). OTIS3 By Neutrophil Presence Across multiple cytological scores, specimens with neutrophils had higher OTIS3 scores than those without (Figure 3a–c). Wilcoxon–Mann–Whitney U tests were performed with FDR correction to adjust for multiple comparisons. For cocci, significantly higher OTIS3 scores were observed when neutrophils were present at cytological scores 0 (q < 0.0001), 1 (q < 0.05), 2 (q < 0.05) and 3 (q < 0.01). A similar pattern was observed for rods and Malassezia, with significantly higher OTIS3 scores at scores 0 (q < 0.0001), 2 (q < 0.01) and 3 (q < 0.001) for rods, and scores 0 (q < 0.001) and 3 (q < 0.0001) for Malassezia. Otitis Index Score 3 (OTIS3) according to cytological scores (0–3) for (a) cocci, (b) rods and (c) Malassezia, grouped by the presence (hatched bars) and absence (white bars) of neutrophils in 318 dogs with otitis externa. At multiple cytological scores, OTIS3 scores were higher when neutrophils were present across all pathogen groups. Comparisons were performed using Wilcoxon–Mann–Whitney U tests with false discovery rate (FDR) correction Significance levels: *, q < 0.05; **, q < 0.01; ***, q < 0.001; ****, q < 0.0001; comparisons without an asterisk are not significant. Discussion This retrospective study indicated that otic cytological findings correlate with clinical severity in canine OE, with patterns differing by pathogen group. Cytological scores for cocci and rods, as well as neutrophil presence, were associated with higher OTIS3 scores. By contrast, Malassezia scores were not consistently related to clinical severity. These results suggest that otic cytological findings provide pathogen‐specific and inflammatory information that complements severity assessment. Bacterial burden was positively correlated with clinical severity, as reflected by higher OTIS3 scores with increased cytological grades for both cocci and rods. This is consistent with previous studies indicating that bacterial infection exacerbates clinical signs of canine OE [4]. Rod‐shaped bacteria, particularly Pseudomonas species, are recognised to cause aggressive, suppurative OE characterised by purulent discharge and marked inflammation [23, 24, 25]. Cocci, most commonly Staphylococcus species, also are frequent isolates and, when present in large numbers, can contribute to erythema, swelling and discharge [23, 26]. Although bacteria in this study were classified only as cocci or rods without species‐level identification, the associations with OTIS3 underscore the clinical value of cytological findings even at a morphological level. By contrast, Malassezia cytological scores did not consistently correspond with OTIS3 scores. This reflects a commonly observed clinical pattern in which the degree of Malassezia proliferation does not always correlate with the severity of inflammation. Malassezia organisms form part of the normal otic microbiota, and their increased numbers are generally considered to reflect dysbiosis rather than a direct driver of clinical severity [26, 27, 28, 29]. Malassezia overgrowth is commonly observed in the initial stages of otitis, while progression to more severe suppurative disease is typically characterised by neutrophilic inflammation and bacterial overgrowth [9, 11, 12]. This is consistent with the concept that shifts in inflammation influence the composition of the otic microbiota and contribute to the development of secondary infection [13]. Accordingly, Malassezia counts may be less reflective of clinical severity than inflammatory responses or bacterial cytological findings. When interpreting these findings across cytological scores, the overall distribution of total cytological scores also should be considered. Most cases were observed within intermediate cytological scores, with relatively few at the extremes (e.g., total scores of 0 and 9). This pattern may reflect underlying microbial dynamics in canine OE [30]. A total cytological score of 9 requires maximal scores for cocci, rods and Malassezia simultaneously. However, severe suppurative otitis is typically dominated by bacterial overgrowth, while Malassezia populations may be less prominent in highly inflammatory environments [9, 31, 32]. Thus, the concentration of more severe cases at intermediate cytological scores may reflect disease progression rather than a simple linear relationship between total microbial burden and clinical severity. This clinical context also is relevant to the referral nature of the study population, as most patients were likely to have undergone prior management. The recorded findings therefore represent the clinical and microbial status at the time of specialist evaluation rather than at the first clinical presentation of the disease. In addition to microbial patterns, neutrophil presence was associated with higher OTIS3 scores in the present study. From a clinical perspective, neutrophil detection in otic cytological findings may help distinguish inflammatory patterns relevant to disease characterisation [11]. In particular, it may help differentiate between erythemato‐ceruminous and suppurative otitis, with direct implications for treatment selection and antimicrobial stewardship [9, 12]. In this context, neutrophil detection may support the use of antimicrobial therapy in cases consistent with suppurative inflammation, while its absence may help avoid unnecessary antimicrobial use in cases of simple microbial overgrowth. Because this study exclusively evaluated intact neutrophils, the degree of inflammation may have been underestimated in certain cases. In severe suppurative otitis, particularly when associated with rod‐shaped bacteria, intact neutrophils are often sparse as a consequence of extensive cell lysis, leaving predominantly neutrophilic debris or nuclear streaming. Incorporating these cytological features may further refine inflammatory assessment in future studies of canine OE. Despite this association, OTIS3 scores overlapped substantially between ears with and without neutrophils, suggesting that neutrophil presence alone did not clearly separate clinical severity. This may reflect that OTIS3 incorporates multiple clinical features of otitis, while cytological assessment represents localised inflammatory findings at the sampling site. The relationship appeared to be more consistent in ears with bacterial pathogens, while differences were less apparent at intermediate Malassezia cytological scores. Notably, neutrophil presence was associated with higher OTIS3 scores even at a cytological score of zero, indicating very low organism numbers rather than absence. This finding suggests that inflammatory responses may be relevant even with minimal microbial burden and that small numbers of organisms may still be clinically meaningful when neutrophils are present. In the present study, cytological evaluation was standardised using explicit count‐based thresholds per oil‐immersion field for cocci, rods and Malassezia, while neutrophils were recorded as present or absent. Although otic cytological assessment was initially considered in the development of the OTIS3 index, it was ultimately not incorporated as a scored element owing to variability in counts and the absence of clear thresholds [10, 12, 20]. By applying predefined criteria, the present study addressed this challenge, allowing more direct comparison between cytological findings and OTIS3 scores at presentation. These findings suggest that otic cytological assessment, using predefined thresholds and neutrophil presence as a binary indicator, could be incorporated alongside OTIS3 to support refinement of the scoring system and improve its applicability in distinguishing clinically relevant inflammatory patterns of canine OE. Given that clinical and cytological assessments in this study were performed by multiple clinicians, the potential impact of interobserver variation should be considered. Previous studies indicate that both the OTIS3 index and semiquantitative cytological grading systems demonstrate good interobserver reliability, supporting their application across multiple evaluators [8, 20]. In our results, a two‐grade difference in cytological grades was generally required to observe significant differences in OTIS3 scores. This pattern may reflect potential scoring overlap between adjacent intermediate tiers under multi‐observer conditions. A similar principle has been recognised in other clinical grading schemes, such as canine mast cell tumour grading, where reducing the number of tiers has been used to improve reproducibility [33]. Therefore, although multi‐tiered systems yield detailed clinical data, compressing the cytological scoring system into fewer tiers could be a practical approach to reduce interobserver variation. Several limitations should be acknowledged. First, this was a retrospective, single‐centre study, which may affect the generalisability of the findings. As cases were recruited from a referral dermatology service, some dogs may have been referred after management in primary care and represent more advanced or chronic cases. This may explain the relatively higher median age observed in this cohort, and the findings should therefore be interpreted in the context of a referral population. Second, selecting the more severely affected ear in bilateral cases also may introduce selection bias. Third, bacterial organisms were categorised morphologically as cocci or rods without species‐level identification, restricting interpretation to broad cytological categories. Fourth, mixed infections were not distinguished, which may have influenced OTIS3 scores within each cytological category. Finally, as cytological scoring thresholds in this study were defined according to institutional protocols without external validation, further prospective studies are needed to establish standardised and reproducible criteria. In conclusion, this study demonstrates that otic cytological assessment and OTIS3 provide complementary information for the assessment of clinical severity in canine OE, with patterns differing by pathogen group and inflammatory status. Integration of cytological evaluation alongside OTIS3 may therefore improve clinical assessment and support a more structured approach to the management of canine OE. Author Contributions Eun‐Seo Shin: conceptualization, data curation, formal analysis, investigation, methodology, writing – original draft. Soh‐Yoon Kang: investigation, methodology, writing – review and editing. Min‐Sun Kim: investigation. Cheol‐Yong Hwang: conceptualization, project administration, supervision, writing – review and editing. Funding The authors have nothing to report. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Sex distribution of 318 dogs diagnosed with otitis externa. Table S2: Breed distribution of 318 dogs diagnosed with otitis externa.