# Pain Assessment in Laboratory Animals: Behavioral and Physiological Indicators


## Key Takeaways

- Pain assessment in laboratory animals necessitates the integration of both behavioral and physiological indicators, as no single sign is diagnostic; reliance on a single parameter, such as body weight or food intake, is a common failure mode.
- Behavioral indicators encompass spontaneous changes in posture, grooming, and activity, alongside provoked responses to stimuli and validated facial grimace scales, which require rigorous training and are best used as part of a composite assessment.
- Physiological indicators, including heart rate variability, neuroendocrine markers (e.g., corticosterone), and body temperature, reflect autonomic and hypothalamic-pituitary-adrenal axis activation but lack pain specificity, responding to various stressors like handling and restraint.
- Composite scoring systems, validated for specific species, strains, and procedures, are crucial for improving inter-observer reliability and enabling longitudinal tracking of pain, with validation requiring correlation to known painful states and response to analgesia.
- Confounders such as handling stress, environmental enrichment, social housing, circadian phase, and prior experience significantly influence pain expression and interpretation, mandating baseline data collection and systematic observation within the animal's specific context.
- Documented pain assessment protocols must specify assessment tools, timing, frequency, observer training, blinding, and predefined action thresholds for analgesic intervention or humane euthanasia, reviewed and approved by institutional animal care and use committees.

---

Pain assessment in laboratory animals is a core component of humane experimental design and a prerequisite for valid scientific data. Unrelieved pain confounds physiological, behavioral, and cognitive endpoints, and it compromises the translational value of research models. This article provides a structured framework for recognizing, scoring, and documenting pain in common laboratory species, with emphasis on behavioral and physiological indicators that can be applied in real time by veterinary staff and research personnel.

The intended reader is a veterinary researcher or laboratory animal clinician who requires a diagnostic reference instead of an introductory overview. The article addresses the following questions: which behaviors are reliable pain indicators across species, which physiological parameters support or refute a pain diagnosis, how scoring systems are constructed and validated, and where the evidence base remains contested. Pain management therapies are excluded from this article, the focus is assessment only.

Accurate pain assessment begins with an understanding of the animal's baseline. Laboratory animals are prey species, and many display evolved suppression of overt pain behavior. The absence of vocalization or escape behavior does not indicate the absence of pain. Assessment therefore requires systematic observation, knowledge of species-typical behavior, and integration of multiple indicators instead of reliance on any single sign.

## At a Glance

| Parameter | Key Information |
|---|---|
| Primary assessment principle | Integrate behavioral and physiological indicators, no single sign is diagnostic |
| Behavioral domains | Spontaneous behavior, provoked responses, facial expression, activity, grooming, feeding |
| Physiological domains | Heart rate, respiratory rate, body temperature, neuroendocrine markers, body weight |
| Scoring approach | Composite scoring systems with defined criteria per species and procedure |
| Major confounders | Handling stress, environmental enrichment, social housing, circadian phase, prior experience |
| Validation requirement | Scores must correlate with known painful states and respond to analgesic intervention |
| Documentation standard | Baseline, scheduled post-procedure intervals, and unscheduled assessments recorded in animal records |
| Regulatory context | Institutional animal care and use oversight per national and international standards |

## The Conceptual Basis of Pain Assessment

Pain is a multidimensional experience comprising sensory-discriminative, affective-motivational, and cognitive-evaluative components. In non-human animals, pain can only be inferred through observable correlates. The inference is strongest when behavioral and physiological measures converge and when the measures respond predictably to analgesic intervention.

The neurobiology of pain involves nociceptive pathways from peripheral receptors through the spinal cord dorsal horn to supraspinal centers, with descending modulatory systems that can amplify or suppress transmission. Physiological pain indicators reflect activation of the autonomic nervous system and the hypothalamic-pituitary-adrenal axis. These responses are not pain-specific, they also occur with fear, restraint, and other stressors. This lack of specificity is the central limitation of physiological measures and the reason behavioral assessment must accompany them.

Behavioral indicators are similarly imperfect. Pain-related behaviors may be suppressed in the presence of observers, masked by competing motivations such as exploration or feeding, or confounded by the experimental procedure itself. Surgical manipulation, for example, produces tissue trauma that alters locomotion independent of pain. The elevated plus-maze literature illustrates the broader problem: ethologically derived behavioral indices can enhance sensitivity to subtle changes in affective state, but the inclusion of large numbers of indicator variables introduces factorial validity problems and interpretive complexity [Wall and Messier, institutional publication on elevated plus-maze methodology](https://pubmed.ncbi.nlm.nih.gov/11378181/). Pain scoring systems face the same risk. A parsimonious set of well-validated indicators outperforms a lengthy checklist of poorly understood signs.

## Behavioral Indicators of Pain

### Spontaneous Behaviors

Spontaneous behaviors are observed without provoking the animal. They include changes in posture, locomotion, grooming, feeding, and social interaction. A hunched posture with abdominal tucking is a classic sign of visceral or abdominal pain in rodents and rabbits. Pilocrection, reduced grooming leading to a dull or stained coat, and reduced food and water intake are common accompaniments. Activity levels may decrease, although some animals show restlessness, pacing, or repetitive movements.

The specificity of these signs is limited. Reduced activity follows any major surgical procedure, regardless of analgesic adequacy, and social isolation or environmental change can suppress grooming independent of pain. Baseline observation is therefore essential. Animals housed in social groups may show withdrawal from conspecifics or increased aggression when painful. Conversely, some species, particularly rabbits, may become immobile and quiet, which can be mistaken for calm.

### Provoked Behaviors

Provoked assessments apply a standardized stimulus and measure the response. Palpation of a surgical site, gentle pressure on a limb, or handling that stretches the body wall can elicit guarding, flinching, vocalization, or attempted escape. These responses are graded on ordinal scales, typically from no response to exaggerated response. The value of provoked assessment lies in its ability to detect localized pain that may not alter spontaneous behavior.

Provoked responses are vulnerable to handling-related stress and to learned associations. An animal that has experienced painful handling may respond to the handler's approach instead of to the stimulus itself. Conversely, animals may habituate to repeated palpation. Standardization of the stimulus, the handler, and the testing environment is required for meaningful comparison across time points.

### Facial Expression Scoring

Facial expression scoring has become a widely used tool for rodents, rabbits, and other species. Grimace scales score action units such as orbital tightening, nose bulge, cheek flattening, ear position, and whisker change. These scales were developed from photographs of animals in known painful states and validated against analgesic response. They offer the advantage of being non-invasive and applicable to single animals in their home cage.

The limitations of grimace scales deserve attention. Scoring requires training and is subject to observer bias, although blinded scoring mitigates this. The scales capture pain at a moment in time and may miss intermittent pain. Some action units, particularly orbital tightening, can occur with non-painful distress. Grimace scales are best used as one component of a composite assessment instead of as a standalone diagnostic.

### Cognitive and Behavioral Testing

Experimental protocols may include cognitive or behavioral tests that are sensitive to pain. Tests of locomotor activity, grip strength, and sensorimotor coordination are commonly used in pain models. In stroke research, the incorporation of functional outcome assessment has advanced the field, but reports of behavioral deficits often conflict, and significant correlation between histology and behavior is reported in few studies [DeVries et al., institutional publication on cognitive and behavioral assessment in experimental stroke](https://pubmed.ncbi.nlm.nih.gov/11445138/). Discrepancies arise from method of lesion induction, duration of occlusion, strain, timing of testing, and laboratory environment.

This evidence has direct relevance to pain assessment. Behavioral tests used to measure pain must be validated for the specific model, species, and procedure. A test that is sensitive to pain in one strain may be insensitive in another. The laboratory environment, including lighting, noise, and handling history, can overwhelm the pain signal. Proper control groups, including sham-operated animals, are necessary to distinguish pain effects from procedural effects.

## Physiological Indicators of Pain

### Autonomic and Neuroendocrine Measures

Pain activates the sympathetic nervous system, producing tachycardia, tachypnea, and elevated blood pressure. These parameters can be measured with implantable telemetry devices, which allow continuous monitoring without handling stress. Heart rate variability is a more informative measure than heart rate alone, as it reflects the balance between sympathetic and parasympathetic tone. Monitoring the autonomic nervous system via heart rate and its variability offers relevant tools to assess emotional states, complementary to classical adrenocortical measures [Boissy et al., institutional publication on assessment of positive emotions in animals](https://pubmed.ncbi.nlm.nih.gov/17428510/).

Corticosterone in rodents, cortisol in other species, and catecholamines are classical stress markers. They rise with pain but also with any aversive or arousing experience. Blood sampling itself induces a stress response, so samples must be obtained rapidly or via indwelling catheters. Fecal glucocorticoid metabolites provide an integrated measure over hours but lack temporal resolution.

### Body Weight and Food Intake

Body weight is a simple, objective, and repeatable measure. Pain typically reduces food and water intake, producing weight loss over 24 to 48 hours. The magnitude of weight loss correlates with the severity of the procedure. However, food or fluid restriction is itself a common research procedure, and animals adapt to restriction with behavioral and physiologic adjustments that minimize additional stress [Rowland, institutional publication on food or fluid restriction in laboratory animals](https://pubmed.ncbi.nlm.nih.gov/17536615/). Weight loss must therefore be interpreted against the animal's expected growth curve and the experimental protocol.

### Body Temperature

Hypothermia follows severe pain, particularly in small rodents with high surface area to volume ratios. Postoperative hypothermia can also result from anesthetic depression and heat loss during surgery. Hyperthermia may occur with acute pain and stress. Temperature measurement via telemetry is reliable, rectal temperature measurement adds handling stress and may itself elevate measured values.

## Confounders and Interpretation

The most common error in pain assessment is over-attribution of physiological changes to pain when they reflect other stressors. Handling, transport, novel environments, and social disruption all activate the same autonomic and neuroendocrine pathways as pain. The reverse error is under-attribution, where pain is dismissed because physiological parameters are within normal range. Chronically painful animals may adapt, showing normalized heart rate and cortisol while behavioral indicators persist.

The experimental context introduces additional confounders. Toxicants that cause neurotoxicity may produce behavioral changes that mimic or mask pain. Acrylamide, for example, interferes with kinesin-related motor proteins, producing distal axonopathy and behavioral changes that can affect reproductive performance in exposed laboratory animals [Exon, institutional publication on the toxicology of acrylamide](https://pubmed.ncbi.nlm.nih.gov/17492525/). A researcher assessing pain in a toxicology study must distinguish treatment-related neurological deficits from pain-related behavioral changes.

Environmental enrichment, social housing, and the animal's prior experience shape pain expression. An enriched environment may increase activity levels, making a pain-related reduction in activity more detectable. Social housing allows observation of withdrawal from conspecifics but complicates individual assessment. The observer must know the animal's baseline behavior in its specific housing context.

## Scoring Systems and Their Limitations

Composite scoring systems assign numerical scores to multiple behavioral and physiological indicators and sum them into a total. The Mouse Grimace Scale, Rabbit Grimace Scale, and various postoperative pain scores for rats are examples. These systems improve inter-observer reliability and allow longitudinal tracking. They are most useful when they have been validated for the species, strain, sex, age, and procedure in question.

Validation requires demonstration that scores are higher in animals with known painful conditions than in controls, that scores decrease after analgesic administration, and that scores correlate with other measures of pain. Many published scales lack full validation. The factorial validity problems identified in ethologically derived behavioral analyzes apply equally to pain scales: adding more indicators does not necessarily improve accuracy and may reduce it [Wall and Messier, institutional publication on elevated plus-maze methodology](https://pubmed.ncbi.nlm.nih.gov/11378181/).

Threshold

## Building a Pain Assessment Protocol

A structured assessment protocol converts isolated observations into a defensible clinical judgment. The protocol must specify which behaviors will be scored, when scoring occurs, who performs the scoring, and how the data are recorded. Without this structure, observer drift and inter-individual variation undermine the reliability of serial comparisons.

### Selecting Assessment Tools

The choice of assessment method depends on the species, the experimental model, the anticipated pain duration, and the available equipment. A three-tiered approach is practical.

| Assessment Tier | Examples | Best Suited For | Limitations |
|---|---|---|---|
| Tier 1: Clinical observation | Posture, activity, grooming, food intake, body weight | All species, minimal handling, daily monitoring | Low sensitivity for mild pain, subjective without scoring |
| Tier 2: Structured scoring | Grimace scales, composite pain scores, provoked response scoring | Rodents, rabbits, postoperative and inflammatory pain | Requires training, species-specific validation |
| Tier 3: Instrumented measures | Telemetry (heart rate, blood pressure), activity monitoring, thermal imaging | Chronic studies, continuous monitoring, refinement studies | Equipment cost, surgical implantation, data volume |

Tier 1 assessment is the minimum standard for any animal undergoing a potentially painful procedure. Tier 2 should be added when the procedure is expected to produce moderate to severe pain or when the experimental endpoint requires sensitive pain detection. Tier 3 is reserved for studies where continuous data are needed or where behavioral observation is confounded by the experimental manipulation itself.

The [Guide for the Care and Use of Laboratory Animals](https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf) requires that institutional programs provide adequate veterinary care, including pain assessment and alleviation. The protocol should therefore be reviewed by the attending veterinarian and the institutional animal care and use committee before study initiation.

### Timing and Frequency of Assessment

Baseline data collection is mandatory. Animals must be observed for at least 24 to 48 hours before the procedure to establish individual normal values for body weight, food intake, activity level, and behavioral repertoire. Group averages are insufficient because individual variation in pain expression is substantial.

Post-procedure assessments should occur at defined intervals. For surgical models, assessments at 1, 4, 8, 12, and 24 hours post-recovery capture the trajectory of acute pain. Thereafter, daily assessment is usually adequate unless the model involves progressive pathology. For chronic models, weekly assessments may suffice, but the protocol must specify criteria for increasing assessment frequency if deterioration is detected.

The timing of assessment relative to the light cycle matters. Nocturnal species such as rats and mice are less active during the light phase, and pain-induced reductions in activity may be masked by the normal diurnal trough. Assessments performed during the dark phase under red illumination detect activity changes more reliably, although this is logistically demanding. Alternatively, automated home-cage monitoring systems can capture activity data continuously across the full circadian cycle.

### Observer Training and Blinding

Observer training is the single most important determinant of assessment reliability. Personnel must be trained to criterion on the specific scoring system used, using video examples and live animals. Inter-observer reliability should be quantified using Cohen's kappa or an equivalent statistic, with a target of at least 0.80 for the primary pain score.

Blinding is essential when pain scores are used as experimental endpoints. The observer should be unaware of the treatment group, the surgical status, or the expected pain trajectory. This is particularly challenging in studies where the intervention itself produces visible effects, such as local inflammation or surgical incisions. In such cases, a second observer who is not involved in the procedure can perform the assessments.

### Documentation and Data Management

Pain scores must be recorded on a standardized form or in an electronic data capture system at the time of assessment. Retrospective recording invites bias and data loss. Each record should include the animal identification, the time and date, the observer, the procedure and postoperative day, the individual indicator scores, and the composite score if applicable.

The protocol must define action thresholds. For example, a composite pain score above a specified cutoff triggers analgesic intervention, a second threshold triggers veterinary examination, and a third threshold triggers humane euthanasia. These thresholds must be defined before the study begins and approved by the institutional animal care and use committee. The [NC3Rs](https://www.nc3rs.org.uk/) provides practical guidance on establishing humane endpoints and refining assessment protocols.

## Species-Specific Considerations

### Rodents

Rats and mice are the most common laboratory species, and their pain assessment relies heavily on facial expression scoring and provoked behavior testing. The Mouse Grimace Scale and Rat Grimace Scale score orbital tightening, nose bulge, cheek bulge, ear position, and whisker change. These scales have been validated for postoperative and inflammatory pain, but they require high-quality images or video and trained observers.

Provoked behaviors in rodents include vocalization, withdrawal responses, and defensive reactions to handling. The assessment of provoked responses must be standardized in terms of stimulus intensity and location. A gentle palpation of the surgical site that produces vocalization or withdrawal is a clinically meaningful finding, but the stimulus must be applied consistently across animals and time points.

Rodents are prey species and mask pain behaviors when observed by humans. The presence of an observer suppresses spontaneous pain behaviors, particularly in mice. Video recording with subsequent analysis, or observation through a one-way mirror, reduces this effect. Automated home-cage monitoring systems that detect changes in locomotion, rearing, and burrowing behavior provide objective data without observer presence.

### Rabbits

Rabbits present specific challenges. They are stoic prey species, and their pain behaviors are subtle. The Rabbit Grimace Scale scores orbital tightening, cheek flattening, nose shape, whisker position, and ear position. Ear position is particularly informative in rabbits, as pain causes the ears to be held back against the body.

Rabbits are obligate nasal breathers, and pain-induced abdominal breathing may be mistaken for respiratory disease. Abdominal pressing, where the rabbit presses its abdomen against the cage floor, is a reliable indicator of visceral pain. Tooth grinding, reduced grooming, and reduced fecal output are additional indicators, but each requires baseline comparison because individual variation is high.

### Fish and Other Non-Mammalian Species

Pain assessment in fish relies on behavioral indicators such as reduced feeding, altered swimming patterns, surface rubbing, and changes in ventilatory rate. These indicators are less well validated than mammalian pain scales, and the evidence base is limited. The [MSD Veterinary Manual](https://www.msdvetmanual.com/) provides species-specific guidance on clinical assessment, but the practitioner must recognize that formal pain scoring systems for fish are not yet established to the same standard as those for mammals.

## Integrating Behavioral and Physiological Data

A single indicator is rarely sufficient to diagnose pain. The integration of multiple indicators improves diagnostic accuracy, but the interpretation must account for the expected effects of the experimental manipulation itself. For example, a surgical procedure that causes tissue trauma will elevate corticosterone levels regardless of whether the animal experiences pain, and an anesthetic protocol may suppress food intake for 24 hours independent of pain.

The assessment protocol should therefore distinguish between indicators that are sensitive to pain and indicators that are specific to pain. Grimace scale scores are relatively specific to pain in rodents, whereas reduced activity is sensitive but nonspecific. A practical approach is to use a composite score that weights specific indicators more heavily than sensitive indicators.

Physiological measures such as heart rate variability and body temperature can corroborate behavioral findings, but they are influenced by handling stress, environmental temperature, and the experimental procedure itself. The [assessment of positive emotional states](https://pubmed.ncbi.nlm.nih.gov/17428510/) through behavioral and physiological measures may provide a complementary perspective, as the absence of pain behaviors does not confirm the absence of pain.

## Common Failure Modes and Corrective Actions

The most common failure in pain assessment is the reliance on a single indicator, particularly body weight or food intake. These measures are sensitive to pain but also to stress, anesthesia, and the experimental procedure. A second common failure is the use of unvalidated scoring systems. The [methodological concerns raised about behavioral test batteries](https://pubmed.ncbi.nlm.nih.gov/11378181/) apply equally to pain scoring: the inclusion of many weakly validated indicators does not improve sensitivity and may reduce reliability.

Observer drift is a third failure mode. Over time, observers may become less stringent in their scoring, particularly when animals are recovering as expected. Regular refresher training and periodic inter-observer reliability checks prevent this drift.

Finally, the failure to define action thresholds before the study begins leads to inconsistent decision-making. When thresholds are defined post hoc, the assessment becomes a justification for decisions already made instead of a tool for decision-making. The protocol must specify, in advance, the score that triggers analgesic intervention, the score that triggers veterinary examination, and the score that triggers humane euthanasia. These thresholds should be based on published validation data where available and reviewed by the attending veterinarian.

## Recognized Complications and Early Detection

Pain assessment protocols fail most often through misclassification of pain as another state, or through delayed recognition of decompensation. The earliest detectable failures are usually behavioral, not physiological. A rodent that stops grooming, a rabbit that ceases cecal coprophagy, or a fish that abandons station-holding all precede measurable changes in heart rate or corticosterone by hours. Detection depends on knowing the baseline repertoire of each animal and on scheduled observation at times when spontaneous behavior is expressed, typically during the dark phase for nocturnal rodents.

Physiological measures fail differently. Autonomic indicators such as heart rate variability reflect arousal, not pain specifically, and they habituate with repeated noxious stimulation. Neuroendocrine measures, particularly glucocorticoid concentrations, rise with any stressor and cannot distinguish pain from handling, transport, or social disruption. Body weight loss is specific but slow, a 10% loss in a rat over 48 hours indicates significant compromise, but by then the window for early intervention has passed. The discriminating check for a suspected false negative is always the same: does the animal respond to a provoked stimulus, and does that response change after analgesic administration?

## Common Errors and Corrective Actions

Less experienced observers make predictable mistakes. The most frequent is scoring behavior without reference to the animal's baseline, so that a normally inactive strain is judged comfortable and a naturally active strain is judged painful. The corrective action is to record baseline observations before any procedure and to use the animal as its own control. A second common error is over-reliance on a single time point. Pain expression fluctuates with circadian rhythm, feeding, and handling, and a single daily score will miss peaks. Scheduled scoring at fixed intervals, with at least one observation during the active phase, is the standard corrective.

A third error is treating facial expression scores as sufficient. Grimace scales have good specificity for moderate to severe pain in rodents and rabbits, but they are less sensitive for mild pain and are unreliable in prey species that suppress expression in the presence of observers. The corrective action is to combine facial scoring with provoked responses and with physiological trending. A fourth error is failing to account for the effects of food or fluid restriction protocols, which can produce behavioral changes that mimic pain, including reduced activity and altered grooming [Rowland's review of food and fluid restriction in laboratory animals](https://pubmed.ncbi.nlm.nih.gov/17536615/). Any assessment protocol must document the restriction schedule and interpret behavior against that baseline.

| Observation | Likely Cause | Discriminating Check |
|---|---|---|
| Reduced activity, normal provoked response | Circadian quiescence, restriction effect, or mild pain | Compare to baseline at same time of day, test provoked response |
| Elevated corticosterone, normal behavior | Handling stress, environmental disturbance, or pain | Repeat after acclimation, check other animals in room |
| Facial grimace present, normal weight | Acute moderate pain | Confirm with provoked response, reassess after analgesia |
| Weight loss >10%, normal facial score | Chronic pain, disease, or restriction | Full clinical examination, review food intake records |
| Aggression on handling | Pain, fear, or learned aversion | Observe from distance, check for wound or swelling |

## Limitations of the Current Evidence

The evidence base for pain assessment in laboratory animals is uneven. Rodent models dominate the literature, and much of what is written about rabbits, fish, and other species is extrapolated instead of directly validated. Behavioral tests developed for one purpose are frequently repurposed without revalidation. The elevated plus-maze, for example, was designed to measure anxiety-like behavior, and its factorial validity is contested even for that purpose, using its component behaviors as pain indicators requires caution [Wall and Messier's methodological review of the elevated plus-maze](https://pubmed.ncbi.nlm.nih.gov/11378181/). Similarly, cognitive and behavioral tests used in stroke research show inconsistent correlation with histological damage, and the same confounders, including method of lesion induction, strain, and testing environment, apply to pain studies [DeVries and colleagues' review of cognitive and behavioral assessment in experimental stroke](https://pubmed.ncbi.nlm.nih.gov/11445138/).

Expert opinion still differs on several points. Whether spontaneous behavior or provoked response should carry more weight in a composite score is unresolved. Whether physiological measures add value beyond behavioral scoring in chronic pain models is also debated, with some groups arguing that autonomic measures capture affective components that behavior misses, and others countering that they add noise without improving discrimination. The role of positive affective states is increasingly recognized as relevant to welfare assessment, but validated tools for measuring positive states in laboratory animals remain limited [Boissy and colleagues' assessment of positive emotions in animals](https://pubmed.ncbi.nlm.nih.gov/17428510/).

## Escalation and Referral Criteria

Escalation is warranted when pain scores rise despite intervention, when an animal fails to resume normal behavior within the expected postoperative window, or when physiological indicators such as body weight or temperature deviate progressively from baseline. In those circumstances, the attending veterinarian should re-examine the animal, review the procedure and analgesic plan, and consider whether the model itself is producing unanticipated pathology. The [Guide for the Care and Use of Laboratory Animals](https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf) requires that institutional veterinary care include prompt treatment of pain and distress, and that animals not meeting humane endpoints be removed from study.

Specialist consultation is appropriate when the pain is refractory to standard approaches, when the species is unfamiliar to the institutional staff, or when the behavioral phenotype is ambiguous. Laboratory involvement may be needed for confirmatory diagnostics, including imaging, clinical pathology, or postmortem examination. Regulatory reporting is required when pain or distress exceeds the limits approved in the protocol, when an animal reaches a humane endpoint without authorization, or when a procedure causes unanticipated suffering. Institutional animal care and use committees and attending veterinarians have overlapping obligations in these cases, and the [NC3Rs guidance on refinement](https://www.nc3rs.org.uk/) provides practical frameworks for adjusting protocols to prevent recurrence.

## Frequently Asked Questions

### How Do I Build a Reliable Pain Assessment Protocol When Budget or Staff Time Is Limited?

Prioritize a small set of validated behavioral indicators over a large, unvalidated checklist. A two-tier approach works well: a brief screening assessment performed at every handling event, and a more detailed scoring session at defined postoperative or post-procedural time points. The screening tier can rely on spontaneous behaviors, facial expression, and body weight trends. The detailed tier adds provoked responses and physiological measures. Restrict the detailed tier to animals that screen positive or that underwent high-pain procedures. The [Guide for the Care and Use of Laboratory Animals](https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf) provides the institutional framework for defining these monitoring standards, and the [NC3Rs resources](https://www.nc3rs.org.uk/) offer practical refinement strategies that reduce cost while preserving data quality.

### What Should I Do When the Species I Am Working With Has No Validated Pain Scoring System?

Use a composite approach built from homologous behaviors and physiological parameters described in closely related species, then validate internally. Record baseline data for each individual before the procedure, because intra-animal comparison is more sensitive than comparison against published norms. Focus on behaviors that are phylogenetically conserved, such as withdrawal responses, guarding, and changes in activity patterns. For fish and other non-mammalian species, rely more heavily on physiological indicators such as feeding suppression and opercular rate, since facial expression scoring is not applicable. Document your scoring criteria explicitly in the study protocol and seek review by the institutional animal care and use committee. The [MSD Veterinary Manual](https://www.msdvetmanual.com/) provides species-specific clinical reference material that can guide adaptation of assessment approaches.

### How Do Food or Fluid Restriction Protocols Interfere With Pain Assessment?

Restriction protocols directly confound several common pain indicators. Body weight loss, reduced food intake, and decreased activity are expected consequences of restriction and cannot be attributed to pain without additional evidence. The physiologic adaptations to restriction, including reduced metabolic rate and altered neuroendocrine output, also affect physiological pain measures. [Rowland's review of food and fluid restriction](https://pubmed.ncbi.nlm.nih.gov/17536615/) documents that animals efficiently reduce energy and fluid losses within 12 to 24 hours of deprivation, which means baseline values for pain indicators shift substantially. If a study requires restriction, establish separate baseline values under the restriction schedule, not under ad libitum conditions. Consider using target weight monitoring and scheduled access periods that align with the animal's nycthemeral rhythms to minimize confounding.

### What Are the Minimum Data I Should Record in the Animal's Pain Assessment Record?

Record the assessment tool used, the raw score for each indicator, the total score, the observer identity, and the time of day. Include the procedure date, the analgesic protocol administered, and the time of last analgesic dose. Note any events that could confound interpretation, such as handling difficulty, environmental disturbances, or concurrent illness. Record body weight at each assessment and indicate whether the animal is on a restriction protocol. The [Guide for the Care and Use of Laboratory Animals](https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf) emphasizes that veterinary care records must document the condition of animals and the treatments provided. A structured form with predefined fields reduces omission errors and supports retrospective review when intervention thresholds are questioned.

### How Should I Communicate Pain Assessment Findings to Investigators Who Are Not Veterinarians?

Frame the discussion around study validity instead of welfare alone. Pain-related physiological stress alters neuroendocrine output, immune function, and behavior, all of which can confound experimental endpoints. [Seegal's review of PCB-induced neurotoxicity](https://pubmed.ncbi.nlm.nih.gov/8958469/) illustrates how behavioral and neurochemical changes interact, and similar logic applies to uncontrolled pain. Present the assessment data as a time series instead of a single score, and explain the intervention threshold in operational terms, such as a defined score that triggers analgesic administration or early euthanasia. Offer to demonstrate the scoring system during a scheduled animal observation session. This approach converts an abstract welfare concern into a concrete data quality issue that investigators can evaluate directly.

### Can Positive Welfare Indicators Be Used to Assess Pain, or Should I Focus Only on Negative Signs?

Positive indicators are useful as recovery markers but are not sufficient as standalone pain measures. The absence of pain is not equivalent to the presence of positive affect, and an animal may show normal exploratory behavior while still experiencing significant pain. [Boissy and colleagues' assessment of positive emotions](https://pubmed.ncbi.nlm.nih.gov/17428510/) describes how behaviors such as play, allogrooming, and anticipatory behavior reflect positive affective states and can complement classical measures. In practice, use the return of species-typical positive behaviors as a recovery milestone after an intervention, but continue monitoring negative indicators until the animal meets all predefined recovery criteria. A composite score that includes both the resolution of negative signs and the reappearance of positive behaviors provides the most complete picture of postoperative recovery.

## Related Clinical & Scientific Guides

* [Refining IACUC Protocols to Minimize Animal Pain and Distress](/knowledge/veterinary-medicine/laboratory-animal-science/refining-iacuc-protocols-minimize-animal-pain-distress)
* [Health Monitoring Programs for Laboratory Animal Facilities](/knowledge/veterinary-medicine/laboratory-animal-science/health-monitoring-programs-for-laboratory-animal-facilities)
* [Anesthetic Risk Assessment in Laboratory Animals: Preoperative Evaluation](/knowledge/veterinary-medicine/laboratory-animal-science/anesthetic-risk-assessment-in-laboratory-animals-preoperative-evaluation)


## References and Further Reading

- [Methodological and conceptual issues in the use of the elevated plus-maze as a psychological measurement instrument of animal anxiety-like behavior.](https://pubmed.ncbi.nlm.nih.gov/11378181/). 2001.
- [Cognitive and behavioral assessment in experimental stroke research: will it prove useful?](https://pubmed.ncbi.nlm.nih.gov/11445138/). 2001.
- [Epidemiological and laboratory evidence of PCB-induced neurotoxicity.](https://pubmed.ncbi.nlm.nih.gov/8958469/). 1996.
- [Food or fluid restriction in common laboratory animals: balancing welfare considerations with scientific inquiry.](https://pubmed.ncbi.nlm.nih.gov/17536615/). 2007.
- [Assessment of positive emotions in animals to improve their welfare.](https://pubmed.ncbi.nlm.nih.gov/17428510/). 2007.
- [A review of the toxicology of acrylamide.](https://pubmed.ncbi.nlm.nih.gov/17492525/). 2006.
- [Guide for the Care and Use of Laboratory Animals, 8th Edition](https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf). National Academies Press, 2011.
- [NC3Rs Resources on Replacement, Reduction and Refinement](https://www.nc3rs.org.uk/). NC3Rs.
- [MSD Veterinary Manual, Professional Edition](https://www.msdvetmanual.com/). MSD Veterinary Manual.

## Related Articles

- [Welfare Assessment Tools for Laboratory Rodents](/knowledge/veterinary-medicine/laboratory-animal-science/welfare-assessment-tools-for-laboratory-rodents)
- [Welfare Assessment of Laboratory Fish: Beyond Zebrafish](/knowledge/veterinary-medicine/laboratory-animal-science/welfare-assessment-of-laboratory-fish-beyond-zebrafish)
- [Anesthetic Risk Assessment in Laboratory Animals: Preoperative Evaluation](/knowledge/veterinary-medicine/laboratory-animal-science/anesthetic-risk-assessment-in-laboratory-animals-preoperative-evaluation)
- [Assessing Thermal Imaging for Pain Detection in Laboratory Animals](/knowledge/veterinary-medicine/laboratory-animal-science/assessing-thermal-imaging-pain-detection-laboratory-animals)
- [Recognizing and Managing Postoperative Pain in Laboratory Rabbits](/knowledge/veterinary-medicine/laboratory-animal-science/recognizing-managing-postoperative-pain-laboratory-rabbits)

> This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.