Effect of owner presence on healing of spontaneous chronic corneal epithelial defects after anterior stromal puncture

Abstract

Objective

To examine the effect of client presence on healing rates of spontaneous chronic corneal epithelial defect (SCCEDs) following debridement and anterior stromal puncture (ASP).

Animals Studied

Sixty‐eight client‐owned dogs.

Procedures

Dogs presenting prior to the COVID‐19 shutdown were assigned to the C group (client in the room, 31/68), while dogs presenting after were assigned to group NC (no client in the room, 37/68). Inclusion criteria were retention of fluorescein, non‐adherent epithelium, persistence for at least 1 week, and recheck within 1 month. Exclusion criteria were concurrent ocular disorders and endocrinopathies. Success was defined as negative fluorescein retention at first recheck. t‐Tests, rank‐sum tests, and chi‐squared or Fisher's exact tests were used to compare findings between groups. Logistic regression was used to determine whether odds of success at first recheck differed between groups or were modified by other characteristics.

Results

Dogs in the NC group were older (9.9 vs. 8.7 years, p = .014) and had more bandage contact lenses (BCLs) placed (65% vs. 29%, p = .003). There were no other significant differences between groups. BCL placement was associated with significantly greater odds of healing by first recheck (OR = 4.00, 95% CI: 0.63–11.2; p = .008). The NC group initially had 2.5 times greater odds of healing than the C group; after adjusting for BCL placement, the association between client location and healing weakened (aOR = 1.80, 95% CI: 0.63–5.13; p = .277).

Conclusions

Healing was marginally associated with not having the client in the room, likely due to increased BCL use. BCL application improves SCCED healing rates following debridement/ASP.

1. INTRODUCTION

Spontaneous chronic corneal epithelial defects (SCCEDs) are superficial corneal ulcers that fail to heal in a normal time frame and are often recurrent. These persistent corneal erosions in dogs are a therapeutic challenge in the clinic setting, evidenced by multiple recommended treatment modalities with varying outcomes for success. Common therapies are cotton‐tipped applicator (CTA) debridement followed by a variety of procedures including anterior stromal puncture (ASP), and grid keratotomy. These procedures manipulate the anterior corneal stroma and non‐adherent epithelium by making small punctures or lines within the affected cornea. ¹ The reported rates of healing in the literature are quite variable with an average of 50% for epithelial debridement alone, and 80% if followed with ASP. ¹ Success rates of debridement with keratotomy vary widely, with reported ranges of 10% (3/29 eyes healed within 14 days), ² 75% (44/52 eyes healed at 10–14 days), ³ 79% (24/30 eyes, median healing of 14 days), ⁴ and 60% (15/25 eyes, healing ranged from 2 to 39 days) ⁵ following a single procedure.

The COVID‐19 pandemic drove an unprecedented shift to veterinary practice in which public health regulations limited public contact with veterinary health care professionals. On March 13, 2020, the University of Wisconsin–Madison took public health‐associated measures and closed its doors to the public to protect the health of clients, staff, and students. Following this shift, examinations and treatments of pets were performed in the absence of their owners. Prior to the pandemic, examinations and treatments, including for SCCEDs, were performed with clients present in the exam room. The practice change during the pandemic allowed us an opportunity to categorize client presence as a variable affecting rates of success in healing for SCCEDs following therapeutic interventions. We wondered if perhaps clinicians (faculty, residents, and students) behaved differently without clients in the room, perhaps by persisting in administering treatments longer even with a recalcitrant patient. Additionally, a new practice model is emerging in veterinary medicine that encourages client presence for most procedures, so examining the effect of client presence on healing rates is increasingly of interest (https://veterinaryemergencygroup.com/our‐story/).

Our study objective was to examine the effect of client presence on healing rates of SCCEDs following corneal debridement with an ASP in awake patients. Our hypothesis was that performing these procedures for treatment of SCCEDs without clients present in the exam room would lead to higher odds of success.

2. MATERIALS AND METHODS

A retrospective case study was carried out by reviewing the medical records of 202 dogs diagnosed with SCCED by an ACVO diplomat or ABVO approved resident at University of Wisconsin Veterinary Care dating from 2015 to 2022. Owners signed a consent form at first admission permitting use of case data and photos for study purposes. All patients received a complete ophthalmic examination including Schirmer tear test 1 (STT), fluorescein staining, and tonometry. Case inclusion criteria consisted of (1) retention of fluorescein staining, (2) evidence of an epithelial defect surrounded by non‐adherent epithelium, (3) no clinical evidence of infection, (4) present for at least 1 week, and a (5) recheck exam within 1 month. Exclusion criteria consisted of (1) systemic endocrinopathies and (2) concurrent ocular disorders such as tear film abnormalities, abnormal lid function/anatomy, uveitis, lens abnormalities, glaucoma, periocular abnormalities, prior ocular surgery, and STT 1 below 15 mm/60 s.

Of the 202 dogs, 68 were eligible for the study after applying the exclusion criteria. Two groups of dogs were defined according to whether the client was in the room during the procedure (group C) or remained outside (group NC), based on their visit date and using March 13, 2020 as the landmark after which clients were no longer permitted in the room. Patients presenting after the COVID‐19 shutdown date were assigned to the NC group (37/68) and were evaluated from May 2020 to April 2022. Canines in the C group (31/68) were patients evaluated from October 2015 to February 2020. Typically in this earlier time frame if a patient had to be handled away from its owner, a notation was made in the medical record, in which case those dogs were excluded.

For the purposes of this study, both groups were followed throughout two appointments. The first appointment consisted of SCCED diagnosis and therapeutic intervention with CTA debridement followed by ASP procedure. If a contact lens was placed, it was a Bausch and Lomb Plano T contact lens (base curve 8.6 mm). Every patient was discharged with (1) prophylactic topical antibiotic therapy, (2) systemic pain management, and (3) an Elizabethan collar. Topical antibiotics used included gentamicin, neomycin/polymyxin/bacitracin, tobramycin, ofloxacin, neomycin/polymyxin/gramicidin, or ciprofloxacin usually given two to three times daily. The second appointment consisted of a recheck evaluation at which success of healing was determined. Success was documented by negative corneal fluorescein staining. Failure was defined as any patient who was not healed at the first recheck.

Signalment was summarized for the overall sample and according to whether the client remained in the room during the procedure (group C vs. group NC). The mean and standard deviation (or median and IQR) were used to describe continuous characteristics while frequencies and percentages were used to describe categorical factors. T‐tests (for age at diagnosis), rank‐sum tests (for weight), and chi‐squared or Fisher's exact tests were used to compare findings between groups. Logistic regression was used to determine whether the odds of healing by the second visit (i.e., at first recheck) were associated with having the client in the room during the initial procedure and whether these odds of healing were further affected by other characteristics (e.g., age, BCL, post‐procedure medication, weight, and affected eye).

3. RESULTS

Sixty‐eight dogs with an average (±SD) age of 9.4 ± 2.2 years at time of diagnosis were part of the study. Approximately half (35/68, 51%) were male and only two (3%) had a prior history of SCCED. In both dogs with a prior history of SCCED, the previous SCCED was in the contralateral eye. All dogs had CTA/ASP as the first procedure and just under half (33/68, 49%) had BCL at the first visit. Forty‐six percent (31/68) of procedures were performed between October 20, 2015 and February 26, 2020 while the client was in the room (group C) and 54% (37/68) of procedures were performed between May 21, 2020 and April 27, 2022 with the client waiting outside the room (group NC).

On average, older dogs were more often treated with the client waiting outside the room (9.9 vs. 8.7 years; p = .014) and the same NC cohort also had a greater number of BCL placements at the first visit (65% vs. 29%; p = .003). Other characteristics appeared not to be associated with whether the client stayed or left the room (Table 1).

TABLE 1.

Distribution of the age, weight, SCCED history, affected eye, bandage contact lens (BCL) placement, and antibiotics administered to dogs treated for SCCEDs with the client in the room and not in the room.

Variable	Full sample (n = 68)	NC (n = 37)	C (n = 31)	p‐value
Age (years) at diagnosis mean (SD)	9.2 (2.2)	9.9 (2.4)	8.7 (1.6)	.014
Male sex, n (%)	35 (51)	16 (43)	19 (61)	.138
Weight (lbs)
n (%) missing	17 (25)	14 (38)	3 (10)	.289
n complete	51	23	28
Median [IQR]	25.0 [11.8,34.9]	23.6 [10.3, 32.7]	25.7 [13.3, 36.5]
History of SCCED, n (%)
No	66 (97)	36 (97)	30 (97)	1
Yes	2 (3)	1 (3)	1 (3)
Affected eye, n (%)
OD	34 (50)	19 (51)	15 (48)	.808
OS	34 (50)	18 (49)	16 (52)
BCL at first visit, n (%)
No	35 (51)	13 (35)	22 (71)	.003
Yes	33 (49)	24 (65)	9 (29)
Post‐procedure medication
Class A	28 (41)	13 (35)	15 (48)	.110
Class B	22 (32)	16 (43)	6 (19)
Class C	18 (26)	8 (22)	10 (32)

Note: t‐Tests (age), rank‐sum tests (weight), and chi‐squared or Fisher's exact test were used to compare findings between groups. Antibiotics were grouped by class into (A) (neomycin/polymyxin/bacitracin, neomycin/polymyxin/gramicidin), (B) (ofloxacin, ciprofloxacin), and (C) (gentamycin, tobramycin).

Abbreviations: C, client in room; NC, client not in room.

The primary outcome, odds of healing rates after one procedure, was not significantly associated with whether the owner was in the room (NC = 68% [25/37] vs. C = 45% [14/31]); Table 2. When the client is not in the room, odds of healing by first recheck are estimated to be 2.5 times the corresponding odds of healing when procedures are performed with the client present (OR = 2.53, 95% CI: 0.95–6.72, p = .063). This effect is confounded by whether a BCL was placed at the first visit. After adjusting for BCL there was less evidence of an association between client location (in vs. out of room) and healing by the first recheck (aOR = 1.80, 95% CI: 0.63–5.13; p = .277). It should be recognized that, by itself, placing a BCL at the first visit is associated with greater odds of healing by the first recheck (OR = 4.00, 95% CI: 1.43–11.2; p = .008), and this effect persists regardless of whether the client was in the room at the time of the procedure (aOR = 2.94, 95% CI: 1.03–8.45; p = .045). A total of 33 out of 68 patients had a BCL placed during their first visit (24/37 NC group, 9/31 C group; Table 3). A total of 10 out of 33 records of patients with BCLs placed stated these were still present during their re‐check examination. For these 10 patients, the recheck examination date averaged 14.2 days and 8 out of 10 cases had resolved. Two of 33 records stated that the BCLs were not present during the recheck exam. A total of 21 out of 33 records did not note if a BCL was or was not present at the recheck exam after being placed at the first exam. Therefore, for 23 of 33, the time frame the BCL was in place was unknown. Other variables noted to differ between groups (age and class of post‐procedure antibiotic; Table 1) were not associated with odds of healing by first recheck once BCL was taken into consideration. The resulting comparison between cohorts is given in Table 2.

TABLE 2.

Comparison of healing rates after anterior stromal puncture with the client in the room and not in the room.

Variable	Total (n = 68)	NC (n = 37)	C (n = 31)	p‐value
Healed by second visit, n (%)
No	29 (43)	12 (32)	17 (55)	.063
Yes	39 (57)	25 (68)	14 (45)

Note: Logistic regression used to determine whether the odds of healing were associated with having the client in the room during the initial procedure and whether these odds of healing were further affected by BCL placement.

Abbreviations: C, client in room; NC, client not in room.

TABLE 3.

Total bandage contact lenses (BCLs) placed and recorded retention rates between groups.

Variable	Full sample (n = 68)	NC (n = 37)	C (n = 31)
Total BCLs placed	33	24	9
BCL present at recheck
Yes	10	8	2
No	2	2	0
No record	21	14	7
Time frame BCL in place if present at recheck (n = average days)	14.2	14.1	21

Abbreviations: C, client in room; NC, client not in room.

Time between the first and second visit ranged between 7 and 34 days (median [IQR] of 14 [14–16] days; one dog did not return for first recheck for 34 days). Lag time (between first and second visit) was not associated (p = .793; Wilcoxon rank‐sum test) with whether the client was in the room (14 [13–17] days) or remained outside during the procedure (14 [14–15] days). When attention is restricted to those n = 39 dogs that had healed by first recheck, there again appeared to be no association (p = .098) between lag time and whether the client remained in the room (16 [14–18] days) or waited outside (14 [14–15] days).

4. DISCUSSION

Treating dogs with SCCED in this study without the client present in the exam room had a marginally higher healing rate. This effect was confounded by BCL placement which significantly improved healing. When comparing our study groups, those in the NC group were older (9.9 vs. 8.7 years, p = .014) and had more bandage contact lenses (BCLs) placed (24/37 vs. 9/31; p = .003). We suspect that BCLs were more likely to be placed when the owner was not in the exam room due to decreased discomfort of clinicians when clients are concerned about their dogs. Additionally, clinicians were likely more comfortable taking extra time to place a contact lens when the owners were not present. A major drawback with this study is that BCL retention at the time of the second visit was infrequently recorded, so it is possible that it was not the BCL placement but rather another variable that led to increased likelihood of healing.

There are several studies that have reported similar results with SCCED therapy and BCL placement. One study found that mean healing was 3 days shorter when a BCL was placed (80/129 healed with a contact retention rate of 62%), which led the authors to conclude that BCL use and retention significantly improve healing times. ⁶ A second study reported that BCL placement resulted in a 100% (10/10 eyes) healing rate; compared to 40% (4/10 eyes) healing rate with no BCL placement. ⁷ The authors' conclusion was also that BCL placement was beneficial for SCCED healing along with decreasing patient discomfort. Placement was also associated with minimal risk and was relatively easy to perform. ⁷ A third study noted that in a group of dogs receiving a grid keratotomy, 16% of patients had a BCL placed and all of those patients healed. ⁴

Given the retrospective nature of this study, several limitations are noted. One limitation is that the doctor performing the procedure was not recorded specifically in the medical record and so a comparison by diplomate or resident cannot be made. It is possible that some animals in the client present group actually did not have the client present in the room but it was not recorded anywhere in the record. Prior to COVID, our practice rarely removed dogs from the room with owners, but it could be that some were in fact treated away from the owner. The fact that we placed more BCLs in dogs during COVID suggests that we did behave differently when the owners were not in the room, as it is unlikely that there was some difference in that many SCCED patients that prompted us to place more BCLs, suggesting that the pre‐COVID cases did have clients present more often.

The conclusion of this study was that healing after one procedure was marginally associated with not having the client in the exam room, likely due to increased BCL use. BCL application improves SCCED healing rates following debridement and ASP procedures.

AUTHOR CONTRIBUTIONS

Natalia L. Rivera‐Viscal: Data curation; methodology; writing – original draft. Ellison Bentley: Conceptualization; data curation; methodology; project administration; writing – original draft; writing – review and editing. Taylor A. Opgenorth: Conceptualization; writing – review and editing. Michael R. Lasarev: Formal analysis; writing – review and editing.

CONFLICT OF INTEREST STATEMENT

No conflict of interest to declare for any authors.

ETHICS STATEMENT

This study complies with the Guidelines for Ethical Research in Veterinary Ophthalmology (GERVO) and is exempt from approval by an ethics committee. Animal owners or owners' representatives provided consent for the treatment provided and for the publication of data and images.

Rivera‐Viscal NL, Bentley E, Opgenorth TA, Lasarev MR. Effect of owner presence on healing of spontaneous chronic corneal epithelial defects after anterior stromal puncture. Vet Ophthalmol. 2025;28:625‐629. doi: 10.1111/vop.13274

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.