Assessing the capillary refill time to identify potential shock has been ingrained in anyone that has taken one of the paediatric alphabet courses (APLS/PALS/PLS etc) – but how useful is it really?
Back to the beginning
As with most things in modern medicine, war breeds progress. In 1947 Henry Beecher and colleagues proposed capillary refill time (CRT) as a means of determining shock state. They allied ‘normal’, ‘definite slowing’ and ‘very sluggish’ refill with ‘no’, ‘slight to moderate’ and ‘severe; shock. This theory was based on just over 100 patients. As it entered common usage, initially for the assessment of trauma, an arbitrary value of normal being less than 2 seconds came into use.
Some basic pathophysiology
The idea of using capillary refill time as a surrogate for measuring shock or dehydration status is underpinned by some basic physiological principles. In shocked or dehydrated states there is a degree of vasoconstriction in the peripheral vasculature. This preserves blood flow to the vital organs – the brain, heart and kidneys – at the expense of less vital tissues. As systemic vascular resistance rises, the skin becomes cool and pale with weakened pulses and so capillary refill time becomes prolonged.
How should it be measured?
If you go back to the basic literature there is little agreement on how CRT should be measured. Some papers detail pressing on the nail bed of the right thumb for five seconds then calculating refill using your pocket watch, whilst others pressed on the sternum and used a digital stopwatch to measure. Very few of us, in practice, get out a watch to measure CRT and, in the current ‘bare below the elbows’ climate, many of us may not be wearing a watch at all.
And why do we press for five seconds, not ten or thirty? Is a longer compression time more accurate? Perhaps the true value in measuring CRT is that it provides a cognitive stop point in the management of a sick child – five seconds to get our breathing under control – before going forward. There is data to suggest that pressing for longer leads to a longer CRT (by up to 1.4 seconds) but these results were inconsistent.
Where should it be measured?
Does it matter if it is measured centrally (cCRT), on the sternum, or peripherally (pCRT)? Common sense would suggest that there should be a discrepancy between the two purely as a factor of distance. This has been borne out by a number of studies.
A systematic review in 2014 by Fleming et al. found that in children over 7 days old the upper limit of normal for CRT is 2 seconds when measured in the hands but 4 seconds when measured on the chest or foot.
And if you look at the data from children over a week old there is a fair degree of heterogeneity with capillary refill time on the chest being longer than we are taught.
The thing we really want to know, the bottom line, is it any good? The test is performed daily by all types of healthcare providers on healthy and sick kids but is it actually worth doing? It’s time to dive a little deeper…
Does the age of child matter?
As we saw above, examiners checking the capillary refill time of neonates (at a variety of anatomical locations including hands, feet, sternum and forehead) found that the upper limit of normal was actually 3 seconds, not the 2 we usually use and that it varied with age.
What about the temperature?
The small studies that have looked for a correlation between core temperature and CRT have found none. Making it of less value in detecting a child with a fever.
However, the temperature of the room does make a difference. Gorelick et al. compared the CRT of a group of healthy children of various ages in a warm room and in a cold one. When measured in the warm environment they had a CRT of 0.85 ± 0.45 seconds compared to 2.39 ± 0.76 seconds. This meant that only 31% of those measured in the cold room would be considered as having a normal CRT. And whilst this study is widely quoted only 32 children were tested and across a broad age range.
What’s the interobserver variability like?
If you take the kappa value for the Gorelick study quoted above then there seems to be only a moderate degree of interobserver reliability with a Κ of 0.54. This is not surprising given that only 16 children were used to derive this value. Other studies have shown the kappa value to be as low 0.14 0 a next to worthless test if performed by different operators.
The kappa coefficient or statistic is a marker of interobserver reliability that takes into account the fact that agreement could take place by chance. With a range of -1 to +1 a value of 0.14 indicates slight agreement and 0.54 moderate agreement. This is a matter of debate amongst statisticians.
Champion, one of the proponents of CRT as a core sign in the assessment of shock in trauma, later on recommended not including its use because it is difficult to assess in darkness. Researchers have found that delays in CRT are harder to detect in artificial light or darker conditions.
So how can we improve this? One way is to more accurately measure the refill time than counting out loud ‘One Mississippi, Two Mississippi’ and use a stopwatch. The same person should also be repeatedly checking the CRT rather than relying on a second observer to do so.
So how accurate is it in sepsis or cardiovascular collapse?
For such a widely used vital sign there is very little evidence of its value. Fleming’s review identified just 4 studied totalling 111 children. There is little to no correlation between CRT and blood pressure in neonates. Intensive care based studies looking at CRT and central markers of poor cardiac outcome showed a positive predictive value of 93-96% and a negative predictive value of 40-50% to detect a ScVO2 of less than 70%. A prolonged CRT ≥ 3 seconds is used as an ‘amber’ warning sign in the NICE traffic light guidelines but as we have seen the guideline does not take account of ambient temperature or anatomic location. A recent study published in Archives of Diseases in Childhood showed odds ratios crossing 1 for both pCRT and cCRT for the detection of serious bacterial infection.
What about CRT as a marker for dehydration?
Most of us have been taught that a prolonged CRT indicates possible dehydration. This assumption comes from data from Saavedra and colleagues who even managed to quantify it. They suggested that a CRT of less than 1.5 seconds indicated a less than 50ml/kg fluid deficit, a CRT of 1.5 to 3 seconds indicated a fluid deficit of between 50 and 100ml/kg deficit and if the CRT was longer than 3 seconds then there was a marked deficit of over 100mls/kg. One must be aware that their average normal CRT was only 0.81 seconds (measured at the nail bed) and involved 32 test subjects only.
Capillary refill time varies with age, ambient and skin temperature, the anatomical site of testing and the amount of pressure applied. If it is to be used it should be done in a standardized way using the same operator for repeat testing in order to eliminate inter-observer variability. Trends are of more value than one off measurements.
If you want more practice in recognizing unwell infants consider registering with www.spottingthesickchild.com.
What is the future of capillary refill time as a vital sign?
In the days of ECMO and REBOA surely we should be better able to assess something like capillary refill time than just pressing on some skin and counting out loud? Perhaps there is some sort of app we could use? (Editors note – please feel free to tell me there is)
High speed digital photography has been used in experimental settings to accurately determine capillary refill rate in children with gastroenteritis. The setting, however, was a small single centre trial and the trial had its flaws.
Perhaps the boffins at MIT have the answer with Eulerian video magnification. If you don’t know what I am talking about then take a look at this brief video demonstration. By analysing imperceptible (to the naked eye) movements in individual pixels of a video image then running them through a fancy algorithm, the most minute changes in state can be identified. This might be the way of the future and way towards wireless, non-contact, real-time assessment of vital signs.
The use of CRT as a measurement of dehydration or cardiovascular collapse leaves a lot to be desired and is ripe for further study. Perhaps if we each performed a standardized measure in each child that comes through the ED with a minor injury or minor illness we would have a better idea of normal in no short order.
This post could not be written without the help of the #ADC_JC from October 2014 ably helmed by Alan Grayson.
Beecher HK, Simeone FA, Burnett CH. The internal state of the severely wounded man on entry to the most forward hospital. Surgery 1947;22:672–81
Gorelick MH, Shaw KN, Baker MD. Effect of ambient temperature on capillary refill in healthy children. Pediatrics. 1993 Nov 1;92(5):699-702. full text
Raju NV, Maisels MJ, Kring E, Schwarz-Warner L. Capillary refill time in the hands and feet of normal newborn infants. Clinical pediatrics. 1999 Apr 1;38(3):139-44.
de Vos-Kerkhof E, Krecinic T, Vergouwe Y, Moll HA, Nijman RG, Oostenbrink R. Comparison of peripheral and central capillary refill time in febrile children presenting to a paediatric emergency department and its utility in identifying children with serious bacterial infection. Archives of Disease in Childhood. 2016 Jun 23
Gorelick MH, Shaw KN, Murphy KO, Baker MD. Effect of fever on capillary refill time. Pediatric emergency care. 1997 Oct 1;13(5):305-7.
Strozik KS, Pieper CH, Roller J. Capillary refill time in newborn babies; normal values. Arch Dis Child. 1997; 76:F193-F196.
Fleming S, Gill P, Jones C, Taylor JA, Van den Bruel A, Heneghan C, Thompson M. Validity and reliability of measurement of capillary refill time in children: a systematic review. Archives of disease in childhood. 2015 Mar 1;100(3):239-49.
Raimer PL, Han YY, Weber MS, Annich GM, Custer JR. A normal capillary refill time of≤ 2 seconds is associated with superior vena cava oxygen saturations of≥ 70%. The Journal of pediatrics. 2011 Jun 30;158(6):968-72. full text
Tibby SM, Hatherill M, Murdoch IA. Capillary refill and core–peripheral temperature gap as indicators of haemodynamic status in paediatric intensive care patients. Archives of disease in childhood. 1999 Feb 1;80(2):163-6.
Schriger DL, Baraff L. Defining normal capillary refill: variation with age, sex, and temperature. Ann Emerg Med 1988;17:932-5.
Saavedra JM, Harris GD, Li S, Finberg L. Capillary refilling (skin turgor) in the assessment of dehydration. Am J Dis Child 1991;145:296-8.
Lobos AT, Menon K. A multidisciplinary survey on capillary refill time: inconsistent performance and interpretation of a common clinical test. Pediatr Crit Care Med 2008;9:386-91
Crook J, Taylor RM. The agreement of fingertip and sternum capillary refill time in children. Archives of disease in childhood. 2013 Feb 9:archdischild-2012.
Brown LH, Prasad NH, Whitley TW. Adverse lighting condition effects on the assessment of capillary refill. The American journal of emergency medicine. 1994 Jan 31;12(1):46-7.