The Glasgow Scale Is a Starting Point, Not a Finish Line
Every post-operative orthopedic patient who comes through the door at Skylos Sports Medicine gets a pain assessment before we touch them. For years the Glasgow Composite Pain Scale Short Form has been the industry benchmark for that initial read, and I will not argue against its value. It is validated, it is reproducible, and it gives me a standardized number I can document across sessions. But after 15 years of working hands-on in canine rehabilitation, I can tell you that the Glasgow scale by itself is not enough to guide a safe, effective rehab program.
The Glasgow Composite Pain Scale measures six behavioral categories and scores patients on a 24-point scale. It was developed and validated specifically for acute postoperative and trauma pain in a clinical hospital setting. When you are a technician making 30-second observational calls in a busy surgical recovery ward, it is exactly what you need. Rehab is a different environment with a different timeline, and my job is to read pain expression across a 45-to-60-minute session where the patient moves, loads, fatigues and recovers in front of me.
What I have learned is that the Glasgow scale captures a snapshot. Post-op pain in orthopedic patients, particularly after TPLO, TTA, femoral head and neck ostectomy or tibial plateau procedures, is dynamic. It shifts within a single session. A dog can score a 3 on the Glasgow at intake and show clear signs of breakthrough pain during weight-bearing exercises 20 minutes later. If I only trusted the intake number, I would push too hard.
Behavioral Markers I Track During Every Rehab Session
The behavioral vocabulary of a dog in pain is wide and often subtle. I document the following markers at intake, mid-session and at discharge for every post-op orthopedic patient. None of these are captured systematically by the Glasgow Short Form.
- Spontaneous vocalization pattern: Not just whether the dog vocalizes, but whether vocalization is anticipatory, meaning it starts before contact, or reactive, meaning it follows contact. Anticipatory vocalization tells me something different about central sensitization than reactive vocalization does.
- Facial tension and orbital tightening: The Canine Rehabilitation Institute and pain researchers working from Brondani et al. have highlighted periorbital tightening as a reliable pain indicator. I look for a squinted, narrowed orbital margin at rest, particularly when the limb is passively ranged.
- Respiratory rate shifts: I take a resting respiratory rate before any hands-on work. A rate above 30 breaths per minute at true rest in the absence of thermoregulatory stress is a flag. I log it and recheck after the session cool-down period.
- Weight redistribution during quiet standing: I watch how the patient distributes load across all four limbs during what should be comfortable, unloaded standing. A TPLO patient four weeks post-op who continually lifts the surgical limb or drops the ipsilateral hip when standing on a flat surface is telling me something the Glasgow score at intake did not.
- Response to therapeutic touch versus neutral contact: I make a deliberate point of touching areas remote from the surgical site first. If the dog tightens, lip curls or head-turns to neutral contact on the neck or dorsal thorax, that systemic sensitization tells me the whole nervous system is wound up, not just the local tissue.
- Attention and cognitive engagement: Pain suppresses curiosity. A dog who would normally track movement, engage with novel objects in the rehab space or respond to cues that have been well-trained will disengage when pain is present. I use engagement level as a proxy for pain load, especially in stoic breeds like Labrador Retrievers and working-line German Shepherds who are bred to mask discomfort.
These markers are not a replacement for validated scales. They are the clinical layer I add on top. Together they give me a richer picture than any single instrument can.
Movement Quality and Gait Signals the Scale Cannot Capture
The Glasgow Composite Pain Scale was not designed to capture biomechanical pain expression, and that is not a criticism of the instrument. It is just outside its scope. In a rehabilitation setting, gait observation is one of the most direct pain assessment tools I have.
For post-operative orthopedic patients I pay specific attention to cadence asymmetry. A dog doing a slow walk on the underwater treadmill who is shortening the swing phase on the surgical limb by more than 15 to 20 percent compared to the contralateral limb is compensating. That compensation has a pain-mediated component in the early post-op window and a motor-control component later. Distinguishing between the two requires me to look at the full picture, including surgical timeline, current analgesic protocol and the session response data I have been tracking across visits.
Trunk deviation during ambulation is another marker I log. A dog who drops the pelvis toward the surgical side on stance phase, or who exhibits a consistent lateral trunk shift away from the loaded limb, is unloading through altered core kinematics. This is visible on the underwater treadmill at low speeds and I consider it pain-relevant until the supervising DVM and I have ruled out other causes.
Transitional movement, specifically the sit-to-stand, stand-to-down and stair negotiation sequences, reveals joint-specific discomfort that level walking may hide. A TPLO patient who hip-hikes dramatically during the sit-to-stand or who sits in a "sloppy" asymmetric posture with the surgical stifle rotated externally is showing me compensatory mechanics that map to stifle discomfort. I document these transitions on video whenever possible so the supervising DVM has a visual record alongside my written notes.
The Limits of Owner-Reported Pain Scores
Most orthopedic rehab programs send owners home with a simple numeric pain scale, typically a 0-to-10 visual analog scale or a modified Brief Pain Inventory adapted for veterinary use. I use owner reports as part of the clinical picture, but I weight them carefully because there are well-documented sources of error I deal with constantly.
The first issue is projection. Owners want their dog to be comfortable. I hear this in session after session: "He seemed fine at home, he was eating normally and wagging his tail." Eating and tail wagging are very low thresholds for "fine." A dog can be experiencing moderate chronic post-operative pain and still eat enthusiastically, still greet familiar people, still perform comfort behaviors in the safe home environment. The absence of obvious distress is not the same as comfort.
The second issue is baseline drift. Owners who have been living with a dog through a multi-week recovery stop noticing subtle changes because they recalibrate to a new normal every few days. A client whose dog has been mildly lame for three weeks often reports no change because they are comparing today to yesterday, not today to pre-surgical baseline.
The third issue is breed and individual temperament bias. Owners of stoic, high-drive working breeds routinely under-report pain. Owners of sensitive, communicative breeds like Vizslas or Border Collies sometimes over-report. Neither group is being dishonest. They are reading their individual dog through the lens of their relationship and their expectations, which is exactly what I would do with my own animals.
I use owner reports primarily for at-home behavior change: sleep quality, position preference, willingness to use stairs, changes in appetite. Those behavioral changes over a 24-hour period capture something my 60-minute session cannot. But I never substitute an owner score for my own clinical assessment at the start of a session.
When I Escalate to the Supervising DVM
As a CCRA working under veterinary supervision, my job is to provide the most detailed, data-rich pain picture I can and to communicate promptly when that picture suggests the current analgesic or management plan is not adequate. I do not adjust analgesic protocols. That is the DVM's scope. My scope is recognition and communication, and I take that boundary seriously.
I escalate same-session when I observe any of the following:
- A Glasgow Short Form score that has increased by 4 or more points from the previous session in the absence of a new procedure or surgical intervention.
- Spontaneous, unprovoked vocalization at rest that was not present at the prior visit.
- Refusal to weight-bear on a limb that was consistently loading at the prior session, without a mechanical explanation like joint effusion I can palpate.
- Behavioral signs of allodynia, meaning the dog reacts with pain behaviors to light, non-painful contact in areas remote from the surgical site.
- Autonomic signs: sustained tachypnea, pupillary dilation, or visible trembling at rest in a thermally neutral environment.
I escalate at end-of-session or via written communication when I see slower-developing trends: three consecutive sessions with creeping Glasgow scores even if still below the acute threshold, progressive reduction in active range of motion that I cannot attribute to scar tissue restriction alone, or consistent owner reports of disrupted sleep and positional discomfort that do not match the expected recovery trajectory for the procedure type.
In my role as a Veterinary Reviewer for the TheraPetic® Healthcare Provider Group, I see documentation from rehabilitation professionals across different practice settings, and one of the most common gaps I observe is underreporting of pain escalation signals to the supervising DVM. The fear of "bothering" the vet is real but it is also a patient safety issue. Clear, documented, data-backed communication is not overstepping. It is the job.
Building a Multimodal Pain Picture Across the Recovery Arc
The recovery arc after major orthopedic surgery typically spans 12 to 20 weeks depending on procedure, patient age, body condition and concurrent pathology. Pain expression changes substantially across that arc and my assessment approach shifts with it.
In weeks one through three, acute nociceptive pain dominates. The Glasgow scale is at its most useful here because the patient is expressing acute pain behaviors the scale was designed to capture. My behavioral and gait overlays are supplemental in this phase.
From week four through week eight, the picture gets more complex. Acute pain should be resolving, but this is where I start watching for maladaptive compensation patterns that can develop their own pain generators. A dog who has been consistently offloading the surgical stifle since surgery will often develop contralateral limb overuse pain, lumbar paraspinal tension and hip flexor tightness that are not captured by a stifle-focused pain assessment. I add soft tissue palpation of the lumbar region, sacroiliac junction and contralateral hip into every session in this window.
After week eight, when patients are typically entering higher-intensity land-based exercise, I am watching primarily for activity-associated pain flares. These often show up not during the session but 12 to 24 hours afterward, which is why my owner communication at this stage is specifically targeted: I ask about the night after and the morning after each session, not just overall function.
The Veterinary Society of Surgical Oncology and organizations like the American College of Veterinary Sports Medicine and Rehabilitation have consistently emphasized multimodal assessment as the standard in post-operative canine pain management. My clinical practice reflects that framework. The Glasgow Composite Pain Scale is in my toolkit every single session. It just is not the only thing in the toolkit.
Systematic, layered pain assessment is how I protect patients, support the supervising DVM's clinical decisions and give owners accurate expectations about where their dog actually is in recovery. A validated scale gives me a number. My 15 years of rehabilitation experience gives me the context to know what that number means in this patient, at this stage, on this day.