Comparison of classic and needle arthroscopy to diagnose canine medial shoulder instability: 31 cases

Abstract

This retrospective study compared surgery time, anesthesia time, and costs recorded with classic arthroscopy or needle arthroscopy when diagnosing canine medial shoulder instability. Signalment, examination findings, diagnostics, anesthesia time, surgery time, treatment, invoices, and complications were reported. All cases (classic arthroscopy, 14 cases; needle arthroscopy, 17 cases) were diagnosed with medial shoulder instability. Anesthesia times, surgery times, and invoices were statistically compared for classic and needle arthroscopy (P < 0.05). No significant differences were reported for surgery time (P = 0.13) but existed for anesthesia time (35 minutes shorter with needle arthroscopy; P < 0.0001) and invoice (38% lower with needle arthroscopy; P < 0.0001). No complications were recorded by the time of last direct follow-up, which was at a mean of 12.4 weeks after surgery. Needle arthroscopy offers an alternative, safe technique to reliably diagnose canine medial shoulder instability. Shorter anesthesia times and lower costs to the client may be advantages of needle over classic arthroscopy.

Introduction

Medial shoulder instability (MSI) is a well-described, common cause of thoracic limb lameness in the dog (1–9). It typically results from separate or combined injuries to the medial shoulder structures, namely the subscapularis muscle tendon, the medial glenohumeral ligament, and the medial joint capsule (1–8,10). Reported imaging techniques to diagnose MSI include radiographs (RADS), ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and classic arthroscopy (CAS) (1–8,11). The latter is considered the “gold standard” to diagnose MSI (5). Classic arthroscopy provides high-resolution images and allows accurate inspection of the joint; if indicated, portals for probing of joint structures and simultaneous treatment of an underlying condition can be performed (12). Possible downsides of CAS can be costs of purchase and maintenance, large amount of storage space, cumbersome size precluding easy transport, need to move patients into the operating theater for surgery, duration for and risk of anesthesia, possible risk of infection or accidental damage of joint structures, and associated expenses. Although those shortcomings are acceptable when a clear indication for arthroscopic treatment using CAS is given (e.g., treatment of osteochondrosis dissecans of the humeral head), in cases in which arthroscopy (AS) is to be used for diagnostic purposes only, other imaging modalities might be preferable (13).

One such alternative diagnostic technique to CAS, possibly not associated with those shortcomings, could be needle arthroscopy (NES) (14–21). Compared to CAS, NES equipment is less expensive, needs less storage space, and the portability allows easy transport (15). The diameter of the NES equipment (1.0 to 1.6 mm with 2.0 to 2.5 mm cannulas) is also smaller than that of the CAS equipment. The smaller size is a reported advantage for humans and horses, in whom/which often 4.0 mm or even larger CAS diameter cameras are often used (16,21). However, whether there is a clinically relevant difference in size compared to the commonly used canine CAS arthroscopes (1.9 to 3.0 mm diameter with 2.4 to 4.3 mm cannulas), is unknown. The 0° to 10° angle semirigid NES arthroscopes of 1 system are passed through similar small cannulas (2 to 2.5 mm) into the joint (Biovision Veterinary Endoscopy, Denver, Colorado, USA) (16). Another system features a 0° camera, providing a 120° field of visualization due to a retractable needle lumen (1.6 mm total size diameter; no separate cannula; Trice Medical, Malvern, Pennsylvania, USA) (22). Another system has been described for dogs (0° telescope; 1-mm arthroscope; 1.3-mm cannula; Karl Storz Veterinary Endoscopy, Goleta, California, USA) (20). Irrigation is achieved by an assistant injecting fluids from a sterile syringe, or the joint is distended before introduction of the NES instrument (15,16); a pressure bag can also be used. The NES instrument can be re-sterilized and re-used (15).

In humans, NES is used to diagnose joint injuries rapidly, accurately and, due to in-office use, inexpensively (17,21). Shoulder NES was validated as safe, with no major complications in 1419 cases (22). A comparative study reported that complete accuracy of all pathologies in 106 human knees was established in 91.5% of cases with NES, in 100% of CAS, and in 61% of cases with MRI (23). Needle arthroscopy was less expensive than CAS or MRI (16,17,21). Based on publications on humans (17,21–24), horses (15,16,19), and abstracts on dogs (14,18,20), NES image quality during the actual procedure was sufficient to establish a diagnosis. That work confirmed the results of earlier peer-reviewed studies in humans, including a direct comparison assessing image quality between CAS and NES, yielding the same diagnostic results (13), supporting statements that in-office NES had sufficient image quality (25). Further technological advancements and improvement of NES image resolution and quality, specifically to diagnose abnormalities in the human glenohumeral joint, have been recently reported in the peer-reviewed literature (26). Finally, based on a canine cadaveric study, when specifically assessing canine medial shoulder disorders, NES provided the same diagnostic value when directly compared to CAS (27). However, the lower resolution of NES images, a result of the small diameter optic (19), was evident when projected on large screens. As this is clinically irrelevant, a direct comparison of image quality of NES to CAS is not warranted. In contrast, there is a lack of information on the duration and costs for clients when comparing CAS with NES in the dog. For the canine shoulder, the reported clinical use of NES is limited to an abstract on 4 cases (18), and, when specifically diagnosing canine MSI, NES has not been described or compared to CAS in a large clinical patient cohort. Indeed, to the best of our knowledge, there is no peer-reviewed publication on canine NES in the veterinary literature.

The purpose of this clinical communication was to demonstrate that NES can be used to diagnose canine MSI. Furthermore, we wanted to compare CAS with NES regarding surgical time, anesthetic time, and final invoice. We hypothesized that those parameters are all shorter/lower with NES.

Materials and methods

Medical records of all dogs undergoing shoulder arthroscopy with the diagnosis of MSI at Friendship Surgical Specialists (Institution A) and Sirius Veterinary Orthopedic Center (Institution B) from January 2017 to January 2020 were reviewed. Written owner consent was obtained for all cases before surgery. The choice of CAS versus NES was made based on surgeon preference and following a discussion with the owners. Owners had been informed that, with CAS, the procedure would be performed in the operating room, a diagnosis could likely be established, and if needed, treatment could be simultaneously performed. In contrast, NES was offered solely as a diagnostic method in the preparation area. Owners had been presented either with cost estimates for both options (Institution A, with estimates for CAS being higher than for NES), or with one estimate providing a wide range (Institution B). The low end of the range was for solely NES, the high end of the range was for a situation in which NES would indicate that moving to the operating theater and performing CAS would be needed. Such needed conversion from NES to CAS was explained to be necessary should the severity of MSI assessed to be too high (Grades 3 or 4) (9) or another intraarticular condition (e.g., a biceps tear requiring CAS biceps tendon release) be diagnosed. Higher costs on the estimate were attributed to the additional time and material needed for CAS. To avoid influence on anesthesia and surgery times from surgical treatment for MSI, only dogs were included for which a conservative treatment was recommended. This recommendation was made following arthroscopic evaluation of the shoulder joint with either arthroscopic technique, and establishment of MSI Grades 1 or 2 (9). Medial shoulder instability Grades 1 or 2 are typically treated conservatively. Patients with thoracic limb pathology other than MSI (such as biceps tears, elbow disease) and cases that had an additional procedure performed (e.g., lipoma removal, castration) were excluded. Based on those criteria, cases were excluded from this study during the review of all medical records, when summarizing data for this current study.

All dogs were routinely prepared for aseptic surgery under anesthesia. The difference between CAS and NES was that CAS involved shaving the entire limb from carpus to dorsal thoracic midline (as this is the standard type of preparation for CAS in our institutions and is based on general guidelines for patient preparations undergoing orthopedic procedures) (28). Needle arthroscopy, however, involved shaving an area ~6 × 6 cm, centered over the tip of the acromion process (similar to previous studies on NES) (15,29,30). Regardless of arthroscopic technique, cases that received an injection of platelet-rich-plasma (PRP) also had a small area on the neck shaved to access the jugular vein (Figure 1). Arthroscopy was performed by 4 Board-certified or residency-trained surgeons (DVP, DT, CH, MG), all of whom had experience from at least 100 shoulder AS procedures. For CAS (2.4-, 2.7-, or 3.0-mm arthroscope with 2.9- to 3.7-mm cannulas; Stryker, Kalamazoo, Michigan, USA; or Arthrex, Naples, Florida, USA), the patient was moved into the operating theater, the surgery team donned surgical hats, masks, and sterile gowns/gloves, and the entire patient was draped, as recommended for sterile procedures in an operating room (28). For NES [1.0-mm diameter arthroscope; 2-mm cannula: Biovision Veterinary Endoscopy, Denver, Colorado, USA; or 1.6-mm diameter (entire unit size): Trice Medical, Malvern, Pennsylvania, USA], the surgeon’s sterile attire was limited to gloves, draping was done using an “eye-drape” (Steris Animal Health, Birmingham, Alabama, USA) or “quarter-drapes” (Animal Hospital Supplies, Flowery Branch, Georgia, USA), and arthroscopy was performed in the surgical preparation area, similar to reported in-office NES in humans (30). Joint distension was achieved by sterile injection of saline by an assistant or with a pressure bag (Ethox Medical, Grand Rapids, Michigan, USA). The AS system was introduced immediately distal to the acromion (12). Regardless of arthroscopy technique used, a needle to serve as egress portal was only inserted if intra-articular hemorrhage or cloudiness of synovial fluid limited visualization of the medial shoulder structures. Shoulder manipulation, specifically abduction, allowed assessment of the medial shoulder structures during either scope procedure. If needed, during NES, a non-sterile assistant was instructed to manipulate the limb during shoulder abduction to maintain sterility of the surgeon’s gloves. In the current study, joint probing was not performed because a diagnosis of MSI was possible without that extra step. Imaging material was collected during the procedures (Figure 2). Once the diagnosis of MSI was made, some patients received an intra-articular joint injection consisting of methylprednisolone acetate (Zoetis, Parsippany, New Jersey, USA), hyaluronic acid (Kineticvet, Lexington, Kentucky, USA; or Anika Therapeutics, Bedford, Massachusetts, USA), or PRP (Companion Animal Health, New Castle, Delaware, USA). The skin was closed with 1 intradermal poliglecaprone 25 suture of 3-0 or 4-0 (Monocryl; Ethicon, Somerville, New Jersey, USA), or a small amount of tissue adhesive (Vetbond; 3M, St. Paul, Minnesota, USA).

Figure 1.

Differences in shaving and draping between classic shoulder arthroscopy (CAS; A,B) and needle arthroscopy (NES; C,D). The entire thoracic limb was shaved for CAS (A), and the patient was moved into the operating theater and completely draped; surgeons wore sterile gowns and gloves (B). For NES, patients remained in the surgical preparation area following shaving of a smaller area, centered over the acromion (box with yellow dotted lines; C). If an injection of platelet-rich-plasma was indicated, fur was shaved from the neck to access the jugular vein to harvest blood (shaved area cranial to the yellow box; C). Draping for NES was limited to an “eye-drape” (D), or “quarter-drapes (Figure 2 A’,B’).

Figure 2.

Imaging of NES procedures. Sterile attire was limited to surgical gloves. The 2 systems used consisted of small diameter arthroscopes with included light source (A,A’), permitting micro-invasive access to the shoulder joint (B,B’), providing images of sufficient quality to diagnose MSI (C = injury to subscapularis tendon of insertion, C’ = dorsal labral tear of the medial glenohumeral ligament).

All dogs received properly fitted shoulder stabilization systems (Dogleggs; York, Pennsylvania, USA). Gradual increase of activity and formal physical rehabilitation was recommended for all patients (31). Peri- and postoperative medications varied (Table 1), based on surgeon preference and patient needs. Following PRP injections, nonsteroidal anti-inflammatory drugs (NSAIDs) were withheld for a minimum of 1 wk.

Table 1.

Anesthesia protocols and postoperative care.

Medications used during anesthesia
Indication	Medication
Premedication	Combination of a sedative [acepromazine, 0.01 to 0.05 mg/kg body weight (BW), IV, or dexmedetomidine (0.005 to 0.1 mg/kg IV)], and hydromorphone (0.1 mg/kg BW, IM); occasionally atropine (0.02 mg/kg BW, IV)
Induction	Propofol (4 to 6 mg/kg BW, IV) to effect
Maintenance	Isoflurane in oxygen
Antibiotic	Cefazolin (10 mg/kg BW, IV) approximately 30 min before skin incision and repeated every 90 min thereafter
Intravenous fluids	Isotonic fluids (5 mL/kg/h, IV)
Postoperative treatment
Anti-inflammatory drugs	Carprofen (2.2 mg/kg BW, PO, q12h) or grapiprant (PO, q24h; dose by weight as per manufacturer’s recommendations)
Analgesic	Gabapentin (5 to 10 mg/kg BW, PO, q8-12h)
Sedative	Trazadone (PO, q12h or q8h; dose by weight as per manufacturer’s recommendations)

Data recorded included signalment, affected limb, cause of injury, time from injury to surgery, previous treatment, lameness score at initial and last examination (32), orthopedic and physical examination findings, imaging-method (RADS, US, CT), surgery time, and anesthesia time. Anesthesia time for imaging and joint injections was not included to eliminate this as a confounding factor. Additional recorded data included arthroscopy system used (CAS/NES), MSI-pathology, and adjusted final invoice. The latter was defined as charges limited to anesthesia and AS. Charges for joint injections were removed because significant price variations existed and they were not performed in all cases. Complications were as previously suggested (33).

Statistical analysis

Descriptive statistics were used when indicated. Data were assessed for normality using the Shapiro-Wilk test. To avoid any breach of institutional confidential data, all figures for invoices were not presented as actual cost in dollars and differences between CAS and NES invoices were given relative to setting the median for all invoices as 100. The statistical analysis of invoice differences was based upon median values with 25th/75th percentile comparisons. Regression analysis was performed when indicated. Wilcoxon rank-sum test was used to calculate P-values; P < 0.05 was considered significant (SAS 9.3; SAS Institute, Cary, North Carolina, USA).

Results

Thirty-one dogs met the inclusion criteria (Tables 2, 3). Further details are available from the authors upon request. Orthopedic examinations suggested shoulder pathology in all cases. Shoulder abduction angles were assessed in 16; all but 2 cases had larger angles on the affected compared to the normal side (Table 3). Shoulder imaging consisted of standard orthogonal RADS in 28 (all including elbows), complete thoracic limb CT in 5, and US in 6. The 3 cases without RADs had CTs. Preoperative imaging ruled out elbow disease in all cases. In all cases, MSI was diagnosed during the arthroscopic procedures (CAS: n = 14; NES: n = 17) (Table 4). There were no significant differences for surgery time (P = 0.13; Table 5), but there were for anesthesia time (shorter with NES; P < 0.0001) and invoice (lower with NES; P < 0.0001). Comparison of invoices, when taking the median of all invoices to be 100, revealed that the median invoice of NES was 38% lower than that of CAS (Table 6). In regression analyses, there were weak relationships between invoice and body weight for CAS (R² = 0.15) and NES (R² = 0.03). Thus, invoices were not significantly influenced by weight. Of 22 intra-articular injections, 2 (both CAS cases) were of hyaluronic acid, 1 (an NES case) of methylprednisolone acetate, and 19 of PRP (12 NES; 7 CAS). There was no significant difference in anesthesia time between cases that did or did not receive PRP (P = 0.6). Median anesthesia times (25th/75th quartiles) were 90 min (66.75/120) for cases without and 94 min (79/110) for cases with PRP injection. Formal physical rehabilitation was performed in 23 cases. Complications were not recorded at last follow-up, which was at a mean (± SD) of 12.4 (± 14.6) wk. Median lameness scores (25th/75th quartiles) were higher at initial presentation [2.5 (2.0/3)] compared to last follow-up [0 (0/1)] (P < 0.0001).

Table 2.

Gender, side, and breed size (number/%) of entire cohort.

Gender
FI	FS	MC	MI	Total
3 (10%)	12 (39%)	15 (48%)	1 (3%)	31 (100%)
Affected side
Left	Right	Total
13 (42%)	18 (58%)	31 (100%)
Breed sizea
Small	Medium	Large	Giant	Total
2 (6.5%)	10 (32%)	17 (55%)	2 (6.5%)	31 (100%)

FI — Female intact; FS — Female spayed; MC — Male castrated; MI — Male intact.

^aSmall = < 15 kg; Medium = 15 to 30 kg; Large = 30 to 50 kg; Giant = > 50 kg.

Table 3.

Patient data (mean ± SD) for the entire cohort.

Patient data	Mean (± SD)
Weight (kg)	30.8 (± 10.7)
Age (mo)	71.7 (± 139)
Initial lameness score	2.5 (± 1.0)
Lameness duration prior to presentation (wk)	28 (± 1157.4)
Shoulder abduction affected side (°)	47.3 (± 7.3)
Shoulder abduction non-affected side (°)	39.4 (± 11.5)
Surgery times (min)	17.2 (± 9.2)
Anesthesia times (min)	96 (± 31.3)
Last direct examination recheck (wk)	12.4 (± 14.6)
Final lameness score	0.4 (± 0.7)

SD — Standard deviation.

Table 4.

Diagnoses, secondary changes, and additional imaging findings for the entire cohort.

Diagnoses
Primary diagnosis (injured structure)				Secondary changes
MGHL	SC	MGHL & SC	Total	Biceps inflammation
10 (32%)	11 (36%)	10 (32%)	31 (100%)	6 (19%)
Additional findings (n = 1 each):
Caudal glenoid grade 1 Outerbridge lesion
Small (2 mm diameter) joint mouse
Humeral head fibrillation
Mineralized supraspinatus

MGHL — Medial glenohumeral ligament; SC — Subscapularis muscle tendon of insertion.

Table 5.

Comparison between classic and needle arthroscopy groups.

Comparison	Weight (kg)	Age (mo)	Surgery time (min)	Anesthesia time (min)	Invoice differenceb
CAS
Median	31.60	70.00	17.50	115.00	1.00
Q1	22.85	41.75	13.00	93	0.45
Q3	37.65	106.00	26.25	136	1.59
NES
Median	28.20	60.00	12.00	80	−0.77
Q1	23.45	36.00	8.50	68	−1.00
Q3	35.35	101.50	19.50	97	−0.38
P-valuea	0.45	0.34	0.10	< 0.0001	< 0.0001

CAS — Classic arthroscopy; NES — Needle arthroscopy; Q1 — 25th percentile; Q3 — 75th percentile.

^aStatistical comparison between CAS and NES.

^bInvoice difference from median of of all invoices, relative to the value for CAS median.

Table 6.

Comparison of CAS and NES invoices taking median of all invoices to be 100.

Measurements	All	Classic (CAS)	Needle (NES)	NES less than CAS (%)
Median	100	141	87	38
Maximum	n.a.	192	101	47
Minimum	n.a.	79	75	5

CAS — Classic arthroscopy; NES — Needle arthroscopy; n.a. — Not available.

Discussion

This is the first report describing the use, experience, and outcome of 31 dogs diagnosed with MSI via evaluation with CAS and NES. A diagnosis of MSI was made on all cases, including all dogs that had NES. Therefore, NES can be used to establish a diagnosis of MSI. When compared to CAS, NES was associated with significantly shorter anesthesia times and costs for the client. Surgery times were not significantly shorter. There were no complications, regardless of arthroscopy technique.

Shorter anesthesia times were attributed to 3 factors: size of the shaved area, transition from preparation area to operating theater with CAS-cases, and the difference in draping and set-up between arthroscopy techniques. Those preparation protocols are standard in our institutions for those procedures. Although not specifically measured nor recorded, the shaved area over the shoulder was smaller for NES and it seems logical that anesthesia times are longer when larger areas are shaved, especially if the difference is a small patch of approximately 6 × 6 cm over the shoulder, compared to shaving the entire thoracic limb from the toes to the thoracic midline (Figure 1). Shorter duration of anesthesia time might be of importance as longer anesthesia is a risk factor for surgical site infection (SSI) (34). However, the incidence of SSI in humans after CAS is 0.01 to 1%; in dogs it is 0.5% (21,35,36). Regardless of AS method, no SSI or complications were recorded in our cases, similar to 1419 human NES procedures with a 0% SSI rate (22). Another reason for extended anesthesia times with CAS cases could be that those were moved into the operating theater and fully draped. In contrast, NES was done in the surgical preparation area where the patient had been induced, shaved, and prepared. In addition, NES required only minimal equipment (Figure 2), demanding less time for setup. It is of note that the differences in shaving and draping were a result of surgeon/facility preference in this study and that others might consider shaving less or using smaller drapes for CAS. Although the mean surgery time for CAS was 6 min longer than for NES, this was not significant and not a clinically relevant difference.

Final invoices were significantly lower with NES than with CAS. However, there is a limitation that, although a statistically different P-value could be calculated, no account could be taken of differences in costings of cases at the 2 institutions (e.g., size/difficulty of handling dogs; location/financial position of institution; different state taxes). Therefore, the value of the statistical analysis for this factor must be cautiously assessed. The basic meaning of the median for NES having been 38% lower than CAS is that, overall, invoices for NES could be less than for CAS. This might be important because NES should allow a reduction in the low-end of an estimate. This could make the difference in a client’s decision to proceed with AS. However, we acknowledge that we as the practice owners can dictate the prices for both CAS and NES, and thus invoice adjustment might be possible. A direct per-use comparison of costs between CAS and NES was not possible due to the different purchase costs of the equipment used. When considering purchase costs, draping, operating room fees, needed equipment, expenses for cleaning, and sterilization, it seems reasonable to assume that the per-use cost of CAS is higher than that of NES. We could address the detected differences in anesthesia time and invoices by doing CAS with minimal shaving, preparation, and draping. A prospective comparative study would be indicated. Should results be similar, there would be motivation to make diagnostic CAS more efficient and quicker. Replacement of anesthesia by sedation for NES in dogs (18) could further reduce costs and procedure times. Sedated NES has been reported in standing horses for evaluation of various joints, including those of the cervical spine (15,16,19). The authors of the current study prefer anesthesia over sedation, as it allows unrestricted, pain-free manipulation of the limb, reduces time pressure and, if needed, permits imaging or preparation for and administration of intraarticular injections.

In the current study, joint probing was not done because a diagnosis of MSI was possible without that extra step. However, if needed, this can be done for either technique. Although not the topic of this report, NES might be economically preferable over CAS as it is less expensive to purchase, only has 1 part to be re-sterilized, and can be re-used (15). Indeed, the authors have used the same NES instrument in over 30 cases. Regardless, the authors note that, in a surgical referral practice, full CAS equipment should be available to treat cases with a clear indication for CAS, or if conversion from NES to CAS is indicated. If arthroscopy is only performed for diagnostic purposes, however, NES might be a good alternative to CAS.

As clinical examination and available diagnostics were highly suggestive of a primary shoulder pathology, and for logistic and financial reasons, many of our cases did not undergo advanced imaging. However, if available and owners can cover associated expenses, this can be considered, primarily to rule out other underlying causes for thoracic limb lameness, including elbow pathology (37). Due to time for imaging and costs, CT might be advantageous over MRI. In addition, and with respect to its sensitivity to diagnose MSI, MRI is also inferior to AS (38), as MRI results in many false findings in human joints (21). However, MRI might be helpful in detecting underlying neuropathy (37), a potential cause for thoracic limb lameness. Shoulder ultrasound, performed after CT, can also be helpful to rule out other soft tissue injury (39). Interestingly, the need for advanced imaging in a study of 1419 humans following NES was only 1.4% (22). Therefore, the diagnostic value considering associated costs with advanced imaging prior to arthroscopy needs to be clearly discussed with owners, particularly if medial shoulder pathology is suspected. Finally, NES can be used if advanced imaging is not available, as it permits assessment of the intra-articular structures and, if indicated, allows administration of intra-articular injections, and helps to design the most appropriate plan for physical rehabilitation.

Although not assessed in this study, another potential advantage of NES with its overall smaller diameter cannula might be that, following an intra-articular injection, the risk of extra-articular leakage through the portal following camera removal might be lower than with CAS. Therefore, it could be hypothesized that the beneficial effect of a targeted intra-articular injection might be higher if performed immediately following NES compared to CAS. However, further study is indicated to confirm such assumption.

Although we did not feel that there is a steep learning curve with NES, previous experience with CAS is helpful to successfully use NES. To standardize cases, we focused solely on dogs with MSI but, similar to CAS (6), other shoulder compartments can be assessed, and pathologies diagnosed via NES. Similarly, elbow pathology can also be detected via NES (14,18). Although we excluded such cases from this study, in clinical cases with concern for elbow pathology, assessment of the elbow joint under the same anesthesia is recommended; simultaneous injuries in shoulder and elbow joints are reported in up to 47% of cases (40). Importantly, we do not suggest NES to be used in lieu of CAS for all cases. The decision on which arthroscopic technique is best used should be based on surgeon recommendation and patient needs. For example, in cases with a clear diagnosis of osteochondrosis dissecans based on radiographs, or in cases highly suspected to suffer from a primary biceps tendon injury, the authors would not recommend NES. If intra-articular imaging seems warranted, however, and is aiming to provide solely a diagnosis, then NES is our preferred choice.

Limitations of this study include its retrospective nature and low case numbers. A prospective randomized study might have revealed different results. Both procedures were not performed on the same patient so there are inter-patient differences that cannot be addressed. As various surgeons were involved, it is possible that surgery times were impacted by each individual. Similarly, anesthesia times might have been influenced by the efficiency of the nursing teams and proper recording (exact start and stop times). The results of our study have to be interpreted cautiously because the detected differences might not have been significant had the same amount of fur been shaved, the same room, drapes, and personal protective equipment been used. However, in our institutions, different protocols are associated with each specific arthroscopy technique. Based on communications with other surgeons, similar protocols are also used elsewhere. Finally, as the surgeons of this report were aware that shoulder abduction measurements are not as helpful (6,10,41,42) as originally described (4), those measurements were not part of each examination. Thus, our cases could not be graded following a previously suggested system (9). However, if using only the reported “typical arthroscopic findings” of that system, all cases would have been categorized as Grades 1–2/4 (9).

In conclusion, NES can be considered an alternative method for diagnosing MSI in dogs. Based on our results, NES is associated with reduced anesthesia times and total client costs compared to CAS. CVJ