Academic Jobs - Home of Higher Ed Logo

wong baker pain scale: the research behind the scale

12 min read0 views

At a glance

The Wong-Baker FACES pain scale was drawn by children in a Tulsa hospital in the early 1980s. Four decades of validation studies, a fifth-vital-sign detour and a copyright fight explain what it can and cannot measure.

brown and beige weighing scale
Photo by Piret Ilver on Unsplash

In 1981 a child life specialist named Connie Baker started work at Hillcrest Medical Center in Tulsa, Oklahoma, and found a problem that had little to do with medicine. Children in pain could not play.

Play is how a child takes the measure of a hospital: it turns a strange room into something you can control for a few minutes. Children who were hurting couldn't do it. And when they did manage to say something about the pain, Baker noticed, they were often not believed. Care slipped.

Donna Wong, a pediatric nurse consultant with a national reputation, was consulting at Hillcrest that year. Baker told her what she was seeing. With the hospital librarian, Peggy Cook, the two women read their way through what existed on pain assessment and came back with a short list: a few adult scales and a few pediatric ones, none of them fit for a five-year-old.

The drawing lesson

Numbers were the fashionable answer even then. The numeric rating scale was spreading, and young children could not use it; the digits meant nothing they could point to. Colour scales failed a different test — children picked colours inconsistently, so a red chosen in one bed meant something else in the next.

Faces worked. So Baker and Wong handed school-aged patients a sheet of paper with six empty circles in a horizontal line, each about the size of a quarter, and asked them to think back to their own pain and draw what it had looked like. More than fifty children took part, from the burn unit and the general pediatric ward. They added hairstyles. They elaborated.

Each set of faces was different, but the eyes, noses and mouths rhymed. A composite of the most frequently drawn features went into a pilot with twenty-five more children. A professional artist drew the final version, with one adult addition: more differentiation in the eyes and eyebrows. The scale was introduced in 1983 and has been reproduced on hospital walls ever since, in English and in translations the foundation does not always control.

Tears, and the argument about the smile

The detail that tells you most about the method is the tears.

Most of the worst-pain faces the children drew included them, even though many of those same children did not cry during painful procedures such as burn dressing changes. Asked why, they were emphatic: you could feel that bad inside. The tears stayed. The instructions now say plainly that a patient does not have to be crying to choose the last face.

The first face is a different matter. It smiles, because children drew a smiling face for no hurt, and the designers kept faith with their source. Clinicians have argued ever since that a neutral anchor would be cleaner.

What the numbers are, and are not

Six hand-drawn faces run left to right, scored 0, 2, 4, 6, 8 and 10. Under them sit words: no hurt, hurts a little bit, hurts a little more, hurts even more, hurts a whole lot, hurts worst. The earliest version was scored from zero to five; the six-step structure was kept because it matched other scales, which made statistical comparison possible. The even-number conversion came later, to sit on the same ten-point metric as adult instruments.

The design works because it offers three routes to one answer. A patient who cannot turn a feeling into an abstract number can point at a mouth. A patient whose first language isn't English reads the eyebrows. The task is recognition, not translation.

What it is not: a measure of mood, comfort or general distress, a use the Wong-Baker FACES Foundation refuses outright. Nor is it a tool for unresponsive patients, or a picture against which a clinician compares the patient's actual face to decide what they feel.

The validation record

The psychometric case is stronger than the scale's detractors allow and narrower than its packaging suggests. Gregory Garra and colleagues at Stony Brook tested it in a suburban pediatric emergency department: 120 children aged 8 to 17 with acute pain rated themselves on the faces and on a 100 mm visual analogue scale, the sliding-line measure known as a VAS. Each successive face corresponded to a VAS about 17 mm higher, and the correlation was 0.90. That is a remarkably tight fit for a six-item ordinal scale.

A follow-up study by the same group, published in Pediatric Emergency Care in 2013, asked the question that mattered: does the scale measure pain, or fear? In 197 children aged 7 to 12, the faces correlated 0.63 with the VAS. Against a 26-item Child Medical Fear Scale the correlation was minus 0.02. Factor analysis separated pain and fear into two clean factors. Among school-aged children in an emergency department, the smiling face was not being read as a mood.

The critique that has lasted came from Christine Chambers and Kenneth Craig, writing in the journal Pain in 1998. They gave 100 children aged 5 to 12 a series of vignettes and randomly assigned them to a neutral-anchor scale or a smiling-anchor scale. Children using the smiling version gave significantly higher pain scores for the no-pain and pain-with-negative-emotion scenarios, and significantly lower scores when a vignette described pain alongside positive emotion. The paper's conclusion was careful and damaging: faces scales with smiling anchors may confound emotional state with pain ratings.

There is a second, quieter problem, about arithmetic. A functional measurement study by Armando Oliveira and colleagues in the Portuguese journal Revista de Enfermagem Referência found sizeable deviations from linearity in children aged 6 to 8, improving in the 9-to-11 group. Scores on the scale should not be treated as interval measures below the age of eight. Two faces apart is not a fixed quantity of pain, and averaged scores in a trial inherit that wobble.

The three-year-old claim

The scale is routinely described — in clinical software, in hospital policy documents, in textbooks — as suitable for ages three and older. That figure is a label, and the evidence behind it is thinner than the label implies.

Carl von Baeyer and colleagues spent a decade on the question, ending with a systematic review published in 2017. They found no published evidence that three-year-olds, without special training, can make reliable use of standard self-report pain scales, and only weak evidence that four-year-olds can. From five upward the evidence is strong.

The literature looked better than it was because of an aggregation habit. In 98 per cent of the hundreds of studies of pain assessment in young children, data for three- and four-year-olds were pooled with data from older children, which flatters the younger ones. Of 182 studies that used a faces scale with young children, exactly one reported age-specific data. In that study the correlation between a child's self-reported face score and an observational score was 0.25 — not statistically significant — at three and four, and 0.80 at five to seven. Up to a third of three- and four-year-olds use only the bottom and top of a scale, treating it as a yes-or-no question. The International Association for the Study of Pain now publishes the practical version of that finding: self-report scales should generally not be used with three-year-olds, though most can communicate that pain is present and where it is.

The proposed fix is almost embarrassingly simple. Ask first whether there is any pain at all. If yes, ask whether it is small, medium or large — a three-way classification that maps onto a 0-10 metric as 3, 7 and 10. It improves validity at four. It does not rescue three.

How a pediatric tool became a fifth vital sign

None of this is what made the faces ubiquitous. What made them ubiquitous was a policy fashion the scale never asked to join.

In 1991 California's legislature passed Assembly Bill 791, requiring licensed health facilities to assess pain whenever vital signs were taken. The American Pain Society ran a pain-as-the-fifth-vital-sign campaign from 1995, and the Veterans Health Administration launched its own initiative in 1999. The Joint Commission's pain standards took effect in 2001 and required accredited hospitals to have policies for assessing and managing pain.

The Commission has spent years correcting the record. It did not invent the fifth vital sign and says it opposes the concept. It never required treatment to zero pain. Its requirement that every patient receive a pain assessment was dropped in 2009 from every program except behavioral health. The standards were rewritten for 2018 around function, non-drug options, prescription monitoring and opioid safety, and the pain questions in the patient survey hospitals were judged on were removed and replaced with questions about pain communication. The Commission's own account reads less like a defence than a warning about how a number becomes a target.

What happens to a number that becomes a target is familiar. Michelle Charette and Gabi Schaffzin, writing in the Canadian Journal of Pain in 2025, trace how efficiency came to sit beside validity and reliability as the mark of an ideal instrument, doing more work than either. Fast scores suit fast throughput. They cite an Oslo ethnography in which triage nurses requested a patient's self-reported number in five of 342 assessments, preferring their own observation. The tool built so that children would be believed had, in that emergency department, become a reason not to ask.

Who owns the faces

There is an archival oddity at the centre of all this. The faces are copyrighted, dated 1983 to the Wong-Baker FACES Foundation, and the name is a registered trademark. Publication, commercial use, research use and most organisational use require permission, and sometimes a licence fee. Colour may not be added: the foundation notes that children chose colours inconsistently in the original research, and treats a coloured copy as an infringement. The scale may not be modified, and may not be repurposed for emotions.

So the instrument's authority rests on an artefact made by children in a Tulsa hospital four decades ago, and that artefact is property. Reproductions circulate anyway, redrawn and recoloured, and pasted into electronic records. For a researcher the practical point is dull but real: build the licence into the study budget and the timeline, because the official file is the one whose reliability was tested. For everyone else it is worth noticing that the thing which makes the scale credible, its origin in children's own drawings, is also the thing that makes it licensable. The foundation's own history is candid about the method and the compromises.

What the evidence supports in practice

The record supports a fairly precise set of habits, and they are less tidy than either the enthusiasts or the sceptics suggest.

  • Self-report first, wherever the patient can self-report. For children of five and over, faces scales are reliable and valid; that is the strongest finding in the field.
  • Teach the scale before it is needed. A child meeting the faces during a painful procedure is being asked to learn and rate at once.
  • Stay with one instrument across a shift and across a stay. Switching between a faces scale and a numeric rating makes the trend line a work of fiction.
  • Pair self-report with observation below school age and for anyone who cannot report. The FLACC scale — Face, Legs, Activity, Cry, Consolability — was built for preverbal and sedated patients and is licensed separately through the University of Michigan.
  • Ask about function alongside intensity. Can you turn over in bed? Can you take a deep breath? The Joint Commission's revised standards push assessment in that direction.
  • Do not chase zero. A smiling first face is a poor goal for a child after abdominal surgery.

Two findings deserve space on the wall next to the scale. In a 2014 audit of 307 children having day-case surgery, published in Pediatric Anesthesia, parents' pain scores matched their children's closely, while nurses and independent observers scored consistently lower — median 0 against the children's median 2. Under-scoring by observers is not a neutral error; it withholds analgesia.

And administration matters as much as the drawing. A 2019 study of 211 children and adolescents presenting with knee pain found the faces scale, given without verbal confirmation, recorded a mean of 1.26 out of 5 — 2.52 on a ten-point scale — against 4.22 on a visual analogue scale, with two-thirds of patients recorded as having no pain while reporting pain on another instrument. A face pointed at quickly in a busy clinic is not a measurement.

What the faces actually did

The scale's achievement was never arithmetic. Before it, a hurting child was a diagnostic problem for adults to solve by inference. After it, the child was a witness. That shift is why six drawings outlasted better-engineered competitors, including the one the field now often prefers: the Faces Pain Scale-Revised, published by Carrie Hicks and Carl von Baeyer in 2001, which replaced the smile and the tears with neutral faces and scored 0 to 10 on a metric that behaves more like a line. That version is cleaner. It has never been as beloved, by children or by their parents.

Every measurement tool eventually gets asked a question it was never designed to answer. The faces were designed so that a child would be believed. Four decades of research since have mostly established the boundaries of that promise: solid from five upward, doubtful at four, and at three closer to a hope than a finding. A smile at the zero end tugs at the numbers, and in children under eight the gaps between the faces are not equal intervals. Nor is a face pointed at in a hurry a verdict on anything.

None of which argues for putting the drawing away. It argues for keeping the instruction attached to it. The point of the scale was never the number. It was the asking, in words a child could answer — and the obligation, once answered, to act.

person standing on white digital bathroom scale

Photo by i yunmai on Unsplash

  • medical-history
  • medical-history
  • wong-baker-faces
  • wong-baker-faces
  • pain-assessment
Portrait of Prof. Clara Voss
About the author

Prof. Clara VossView author

Academic Jobs In House Author

Prof. Clara Voss is a Contributing Writer at AcademicJobs, specializing in humanities and social sciences research. As a Professor of History at Humboldt University of Berlin with more than 20 years in academia, she exa…

Discussion

Sort by:

Be the first to comment on this article!

You

You’ll be asked to sign in before your comment is posted.

Browse by Faculty

Browse by Subject

Frequently Asked Questions

🩺What is the Wong-Baker FACES Pain Rating Scale?

It is a self-report tool for rating the intensity of physical pain, made up of six hand-drawn faces arranged left to right and scored 0, 2, 4, 6, 8 and 10, with short word descriptions beneath each one. A patient points to the face that best shows how much they hurt. It was designed for children and is now used with adults who find number-only scales hard to apply.

📜Who created the scale, and when?

Connie Baker, then a child life specialist at Hillcrest Medical Center in Tulsa, Oklahoma, and Donna Wong, a pediatric nurse consultant working at the hospital, developed it together. Work began in 1981, when more than fifty children on the burn and general pediatric units were given paper with six empty circles and asked to draw how they felt at different levels of pain. The finished scale was introduced in 1983 and is copyrighted to the Wong-Baker FACES Foundation.

🧒What ages is the scale suitable for?

It is described as suitable for ages three and older, but the research does not fully support that floor. A systematic review led by Carl von Baeyer found strong evidence of reliability and validity from age five upward, weak evidence at four, and no published evidence that untrained three-year-olds can produce a graded self-report. Most three-year-olds can still communicate that pain is present and where it is.

🔢How is the scale scored?

Each face carries a number: 0 for the smiling no-hurt face, then 2, 4, 6, 8 and 10 for the crying worst-hurt face. The original version was scored from zero to five, and the six-face structure was kept because it matched other paediatric scales. The even-number version was introduced so results could sit on the same ten-point metric as adult pain instruments.

💧Why are there tears on the last face?

Because the children put them there. Most of the worst-pain drawings included tears, even from children who did not cry during painful procedures such as burn dressing changes. They insisted tears were necessary because you could feel that bad inside. The published instructions acknowledge this: a patient does not need to be crying to choose the last face.

🙂Why does the first face smile, and does that bias the score?

It smiles because children drew a smiling face for no hurt. Research suggests the choice carries a cost. In a study by Christine Chambers and Kenneth Craig published in Pain in 1998, children using a smiling-anchor scale gave higher pain ratings for scenarios describing no pain or pain with negative emotion than children using a neutral-anchor scale, and lower ratings when pain was paired with positive emotion. That is why the Faces Pain Scale-Revised uses neutral faces.

🔒Is the Wong-Baker scale free to use?

Not always. The faces and the name are copyrighted and trademarked by the Wong-Baker FACES Foundation, which handles requests for publishing, commercial, research and organisational use and may require a licensing agreement and fee. The foundation also refuses permission for colour additions, modifications, or use as a measure of mood or emotion. Check the current terms before printing it or building it into an electronic record.

⚖️How does it differ from the Faces Pain Scale-Revised?

The Faces Pain Scale-Revised, published by Carrie Hicks and Carl von Baeyer in 2001, uses six neutral faces with no smile and no tears and is scored 0 to 10. It shows a close linear relationship with visual analogue scales across ages four to sixteen and is copyright of the International Association for the Study of Pain, free for non-commercial clinical and research use. The Wong-Baker scale keeps the smile and tears because children drew them, and is more often preferred by children and parents.

🧑Can adults use the Wong-Baker FACES scale?

Yes. The foundation recommends it for people aged three and older rather than for children only. It is useful for adults who struggle to convert a sensation into a number, including patients whose first language differs from the clinician's, since the faces can be read without verbal instruction in that language. It does not work for patients with severe cognitive impairment or for anyone unable to understand the task.

🚫Can it be used with an unresponsive or sedated patient?

No. The scale is a self-assessment tool and requires the patient to understand the instructions and select a face. It should never be used with an unresponsive patient, and a clinician must not compare a patient's actual facial expression to the drawings to infer a score. For preverbal, sedated or cognitively impaired patients, observational tools such as FLACC or the COMFORT scales are the appropriate instruments.

📊How does it compare with a 0-10 numeric rating scale?

The numeric rating scale asks a patient to attach a digit to their pain; the faces scale asks them to recognise an expression. Children and parents tend to prefer the faces, and children under seven often cannot use a number or a bare line reliably. For older children and most adults, the two instruments correlate well enough to be used interchangeably within a single episode, but mixing them across a hospital stay makes trend data unreliable.

👶What should a clinician do for a three-year-old who may be in pain?

Ask a binary question first — is there any hurt at all — and if the answer is yes, ask whether it is small, medium or large. Pair the child's answer with structured observation and with a parent's assessment, since parents' scores track children's closely while clinicians' tend to run lower. Treat, then reassess. The presence and location of pain can be established at three; a graded intensity score usually cannot.