How to Perform and Document a Capillary Refill Test

A capillary refill test measures how quickly blood returns to a small area of skin after you press on it hard enough to push the blood out. The technique itself is simple: apply firm pressure to a fingertip or nail bed for a set number of seconds, release, and count how long it takes for the blanched (white) area to return to its normal color. But that simplicity is deceptive. The test’s accuracy depends heavily on how you perform it, where on the body you press, the surrounding temperature, and even who is watching the clock. Getting it right and recording it properly matters because clinicians use it to make real decisions about perfusion, dehydration, and shock.

The Standard Technique Step by Step

The most widely taught method uses the fingertip or nail bed of the index or middle finger. You hold the patient’s hand at roughly heart level, press firmly on the fingertip pad or the nail bed with your own thumb or finger for about five seconds, then release. The moment you let go, you start counting. The time it takes for normal color to return is the capillary refill time, or CRT. In healthy adults at a comfortable room temperature, most readings fall somewhere between about one and two seconds. A pilot study testing standardized pressures found that the large majority of CRT values in healthy volunteers landed between 0.8 and 2 seconds, with all readings staying below 4 seconds.¹

The traditional cutoff for “normal” is two seconds or less in adults and older children. A result above two seconds is generally flagged as prolonged, suggesting the blood is not flowing back into the tiny capillaries as quickly as expected. In critical care, a three-second threshold is sometimes used instead, particularly in sepsis management. But these cutoffs are rough guides, not hard boundaries, and several factors can shift them.

Pressure matters more than most people realize. Pressing too lightly may not fully blanch the capillary bed, giving you a falsely fast result. Pressing too hard or for too long can compress deeper tissue and delay the return. Five seconds of firm pressure is the convention in most clinical guidelines, though some protocols use three seconds. A four-second press is used in some veterinary settings and has been studied for its consistency.² Whatever duration you choose, the key is consistency: use the same pressure and the same hold time every time you test.

Where on the Body to Test

The fingertip is the default site for adults, but capillary refill time varies significantly depending on where you press. CRT is known to differ with the anatomical site of measurement.³ In newborns, for example, average CRT in the hand is a bit shorter than in the foot: one study of healthy newborns found a mean of about 4.2 seconds in the hand and about 4.6 seconds in the foot.⁴ Those numbers are noticeably longer than the two-second adult cutoff, which is why applying a single threshold across all age groups causes problems.

Central sites like the sternum or forehead are sometimes preferred in trauma and critical care, because peripheral sites (fingers, toes) are more affected by cold environments and vasoconstriction. A large secondary analysis of trauma patients found that central CRT (measured at the sternum) was associated with mortality even in patients whose heart rate and blood pressure looked normal, suggesting it may pick up hidden hypoperfusion that peripheral vital signs miss.⁵ The practical takeaway: if you are doing a quick bedside assessment and the patient’s hands are cold, the sternum or kneecap may give you a more reliable read than the fingertip.

Temperature Changes Everything

Ambient temperature is the single biggest confounder of capillary refill time. Cold fingers refill slowly even in perfectly healthy people. One study comparing healthy volunteers in cold versus room-temperature conditions found that the cold group had a mean fingertip CRT of about 4.7 seconds, compared to roughly 2 seconds at room temperature and 2 seconds again after rewarming.⁶ Lower fingertip temperature, not age, was the dominant factor driving slower refill.

The effect is dramatic in children too. A study of healthy kids tested in a warm room (about 26°C) versus a cool room (about 19°C) found that the mean CRT nearly tripled, going from about 0.85 seconds in the warm room to about 2.4 seconds in the cool room. Every single child had a “normal” CRT in the warm room, but only about a third were classified as normal in the cool room.⁷ This means a perfectly healthy child assessed in a slightly chilly emergency department could look like they have impaired circulation.

If you are testing in an environment you cannot control, make a note of it. A CRT of three seconds recorded in a cold ambulance bay carries a very different meaning from the same number recorded in a warm exam room. Documenting the ambient conditions alongside the CRT value is one of the simplest ways to improve the test’s usefulness.

Skin Color and the Visibility Problem

The test relies on watching a color change: skin goes from blanched to its normal shade. In people with lighter skin, this transition is easy to see. In people with darker skin, the color change can be subtle or nearly invisible, which makes the test harder to perform accurately and less repeatable. Research has confirmed that repeatability and reproducibility of CRT measurements are limited for individuals with darker skin.⁸

This is not just an inconvenience. It is a potential source of health disparity if clinicians either skip the test or misinterpret it in patients with darker skin tones. Some workarounds exist: pressing on the nail bed rather than the finger pad makes it easier to see blanching regardless of pigmentation, and using a penlight can help illuminate the color change. Technological solutions are also being explored, including optical devices that measure CRT based on reflected light intensity rather than visual observation, which has shown good performance across all skin types.⁸ Until those tools are widely available, the best practice is to acknowledge the limitation in your documentation and supplement CRT with other perfusion indicators like skin mottling or lactate levels when visual assessment is difficult.

Why Two Clinicians Often Get Different Numbers

CRT has a well-known reliability problem. When two different clinicians test the same patient, they often disagree. A study of nurses and nurse assistants found only moderate agreement on CRT values between observers, with an interclass correlation of 0.62 and a kappa of 0.56 for classifying results as normal or abnormal.⁹ Another study in adults reported even weaker agreement, with a kappa of just 0.38 and limits of agreement stretching nearly two seconds in either direction.¹⁰

The gap between observers comes from several sources: people press with different amounts of force, count differently, and perceive color changes differently. Standardizing the method helps. One study compared an unstandardized visual method with a more structured approach and found that the unstructured method showed essentially zero correlation with objective image-based measurement, while the standardized visual method correlated well. Experienced clinicians also performed better when using video review.¹¹ The lesson is clear: the more consistently you perform and document the test, the more useful the result becomes, especially when someone else needs to compare it to your reading later.

How to Document It Properly

A CRT value recorded as just “normal” or “prolonged” is not particularly useful to the next clinician. Good documentation includes several elements that let someone else interpret and compare your finding:

Numeric value: Record in seconds, not just as “normal” or “delayed.” “CRT 3 seconds” is far more useful than “CRT prolonged.”
Site tested: Fingertip, nail bed, sternum, toe, or forehead. The location matters because results differ by site.
Ambient conditions: Note if the room was cold, if the patient was outdoors, or if the extremity felt cool to the touch. A brief note like “tested in cool resuscitation bay” adds critical context.
Pressure and duration: Specify if you used a standard five-second press or a different duration. This is especially important in research settings or when tracking CRT serially.
Skin color observation: If visual assessment was difficult due to skin pigmentation or lighting conditions, note that as well.
Time of assessment: Particularly in serial monitoring during resuscitation, recording when each CRT was taken lets the team track trends.

Trending CRT over time is more informative than any single reading. In sepsis management, clinicians may check CRT repeatedly to see whether a patient is responding to fluid resuscitation. A CRT that drops from four seconds to under two over the course of an hour tells a much clearer story than a single measurement in isolation. Documentation should make those trends visible.

CRT in Pediatric Assessment

Capillary refill time has been used for decades as a screening tool in children, particularly for spotting dehydration and serious illness. A systematic review of its diagnostic value in pediatric emergency departments found that CRT had high specificity for identifying meaningful dehydration in children with vomiting and diarrhea, in the range of 89 to 94%. But sensitivity was all over the map, ranging from 0% to 94% depending on the study.¹² In plain terms: a normal CRT in a child is somewhat reassuring, but a normal result does not rule out dehydration.

In pediatric sepsis, CRT thresholds tend to be interpreted in tiers. A CRT under two seconds is generally reassuring and may help rule out a serious condition. Between two and three seconds warrants closer monitoring and further evaluation. Above three seconds is associated with worse outcomes across all ages in patients with sepsis.¹³ For neonates specifically, the picture is murkier. Healthy newborns can have CRTs well above two seconds as a baseline, which makes that standard cutoff essentially useless in the first days of life.⁴

Guiding Resuscitation in Septic Shock

The most significant recent development in CRT’s clinical relevance comes from sepsis research. Rather than using CRT only as a snapshot assessment, recent trials have tested whether targeting CRT as a resuscitation goal changes patient outcomes. The ANDROMEDA-SHOCK-2 trial randomized over 1,400 patients with early septic shock to either a personalized resuscitation protocol guided by CRT or usual care. The CRT-targeted group had better outcomes on a composite measure of mortality, time on life support, and hospital stay.¹⁴ The benefit was primarily driven by shorter duration of vital support in the CRT-guided group.

An earlier trial comparing CRT-targeted to lactate-targeted fluid resuscitation found that using CRT as the endpoint did not lead to less fluid being given, but patients reached their resuscitation target faster. Stopping fluids once CRT fell to three seconds or below appeared safe in terms of tissue perfusion and microcirculatory flow.¹⁵ These findings have pushed CRT from a simple screening tool into something closer to a therapeutic target in intensive care. If you are documenting CRT in a critical care setting, this context matters: the number you write down may directly influence how aggressively the team manages fluids and vasopressors.

Combining CRT with Other Perfusion Signs

CRT works best as part of a larger picture rather than a standalone test. Skin mottling, for instance, is another bedside perfusion marker. In septic shock patients assessed pre-hospital, both skin mottling score and CRT measured before hospital arrival were independently associated with mortality. A CRT above four seconds carried roughly twice the risk of death after adjusting for other factors.¹⁶ The combination of CRT with mottling and other clinical signs gives a richer assessment than any one test alone.

CRT also has some value in vascular assessment, but its limits are real. A study testing whether CRT could identify patients with peripheral arterial disease or impaired lower extremity healing found that it could not reliably discriminate between those groups.¹⁷ In other words, a normal CRT in the feet does not mean the arteries are healthy, and a prolonged one does not confirm disease. For vascular work, CRT is a supplement to more definitive testing, not a substitute.

When CRT Goes in the Wrong Direction

One of the more counterintuitive recent findings about capillary refill time involves blood loss. You would expect CRT to get longer as someone loses blood, since less blood should mean slower refill. But a controlled study simulating progressive blood loss in healthy volunteers found the opposite: finger CRT actually decreased in the majority of participants as blood loss increased.¹⁸ This paradoxical shortening likely reflects early compensatory responses like catecholamine release that temporarily increase peripheral blood flow before the body decompensates.

This finding is worth knowing because it means that in early hemorrhagic shock, a reassuringly normal or even brisk CRT might be misleading. The body may be actively compensating for blood loss in ways that mask the problem at the fingertip. It is another reason why CRT should never be used in isolation, especially in trauma. A patient with a fast CRT but tachycardia, anxiety, and a mechanism suggestive of blood loss still needs aggressive evaluation. The physiology underlying CRT is more complex than “less blood equals slower refill.” During fever, for example, CRT first lengthens during the vasoconstriction phase and then shortens during the vasodilation phase of fever resolution, creating a biphasic pattern that tracks with temperature regulation rather than with any change in the patient’s actual circulatory status.¹⁹

Automated and Optical Measurement Devices

The reliability problems with visual CRT assessment have driven interest in devices that can measure refill time objectively. Prototypes already exist that combine standardized pressure application (via a pneumatic actuator), light-based measurement of blood return (using multi-wavelength reflectance), and temperature sensing, all in one adhesive fingertip sensor.²⁰ These systems remove the two biggest sources of variability: inconsistent pressure and subjective color judgment.

Simpler camera-based approaches are also under development. One method uses a standard RGB video camera with polarized white LED lighting and calculates CRT from the green channel of the video signal. In testing, about 80% of readings fell within 20% of the expected CRT value, with a repetition standard deviation of 17%.⁸ While these technologies are still largely in the prototype and pilot-study stage, they point toward a future where CRT could be measured remotely through a smartphone camera or integrated into continuous monitoring systems. Automated optical determination has been assessed as having high clinical potential, though developers acknowledge that more extensive studies are needed before widespread adoption.²¹

For now, the bedside fingertip press remains the standard. But awareness of these devices matters for documentation: if your facility has access to an automated CRT device, recording whether a measurement was taken visually or by instrument makes comparison over time far more meaningful.

1
Heliyon. Pilot study on optimizing pressure for standardized capillary refill time measurement
2
PubMed. Standardized capillary refill time and relation to clinical parameters in hospitalized dogs
3
Advances in Neonatal Care. CAPILLARY REFILL TIME IS AN UNRELIABLE INDICATOR OF CARDIOVASCULAR STATUS IN TERM NEONATES
4
PubMed. Capillary refill time in the hands and feet of normal newborn infants
5
PubMed Central. Association of central capillary refill time with mortality in adult trauma patients: a secondary analysis of the crash-2 randomised controlled trial data
6
PubMed. Low temperature increases capillary blood refill time following mechanical fingertip compression of healthy volunteers: prospective cohort study
7
Pediatrics. Effect of Ambient Temperature on Capillary Refill in Healthy Children
8
PubMed. Skin-color-independent robust assessment of capillary refill time
9
PubMed. Capillary refill time: a study of interobserver reliability among nurses and nurse assistants
10
Hong Kong Journal of Emergency Medicine. Capillary Refill Time in Adults Has Poor Inter-Observer Agreement
11
The American Journal of Emergency Medicine. The standardized method and clinical experience may improve the reliability of visually assessed capillary refill time
12
PubMed Central. The Diagnostic Value of Capillary Refill Time for Detecting Serious Illness in Children: A Systematic Review and Meta-Analysis
13
Frontiers in Pediatrics. Capillary refill time in sepsis: A useful and easily accessible tool for evaluating perfusion in children
14
PubMed. Personalized Hemodynamic Resuscitation Targeting Capillary Refill Time in Early Septic Shock: The ANDROMEDA-SHOCK-2 Randomized Clinical Trial
15
PubMed Central. Effects of capillary refill time-vs. lactate-targeted fluid resuscitation on regional, microcirculatory and hypoxia-related perfusion parameters in septic shock: a randomized controlled trial
16
PubMed Central. Skin mottling score and capillary refill time to assess mortality of septic shock since pre-hospital setting
17
The Foot. An evaluation of the reliability and validity of capillary refill time test
18
PubMed Central. Capillary refill time paradoxically decreases in a blood loss shock model
19
PubMed Central. Physiological limits of capillary refill time: a human endotoxemia study
20
PubMed Central. An automated quasi-continuous capillary refill timing device
21
Biophotonics Congress 2021. Reliability of automated optical determination of capillary refill time