Silverman Anderson Method for Newborn Respiratory Status

The Silverman Anderson score is a bedside tool that quantifies how hard a newborn is working to breathe, using five visual signs scored on a scale from 0 to 10. Designed in 1956, it remains one of the most widely used respiratory distress assessments in neonatal care worldwide, especially where advanced monitoring equipment is scarce. Its value lies in the fact that a trained observer needs nothing more than their eyes and ears to generate a number that guides real treatment decisions.

What the Score Measures

The Silverman Anderson index evaluates five observable signs of respiratory effort in a newborn. Each item receives 0, 1, or 2 points, where zero means no abnormal sign is present, one means moderate respiratory effort, and two means maximum effort. The total ranges from 0 (no distress) to 10 (severe distress).¹ The five items are:

• Upper chest movements: whether the chest and abdomen rise together normally or see-saw against each other, with the belly pushing out while the chest caves in.
• Lower chest movements: inward pulling of the lower ribs during breathing, also called subcostal retractions.
• Xiphoid retractions: visible sinking at the bottom of the breastbone with each breath.
• Nasal flaring: widening of the nostrils during inspiration.
• Expiratory grunting: an audible sound the baby makes while breathing out.

A score of 0 to 3 is generally interpreted as mild or no significant distress. Scores of 4 to 6 suggest moderate respiratory difficulty, and scores of 7 or above indicate severe distress that typically demands urgent intervention. In practice, many clinicians treat a score of 5 as a key threshold: babies at or above that level are much more likely to need their breathing support increased.

Why These Signs Appear When a Newborn Struggles to Breathe

Each component of the score reflects something specific happening in the baby’s respiratory system. Respiratory distress in newborns commonly presents as rapid breathing, nasal flaring, retractions, and grunting, and it can progress to respiratory failure if not recognized and managed promptly.²

When a newborn’s lungs are stiff or partially collapsed, the diaphragm has to pull harder to draw air in. That extra force creates negative pressure inside the chest, which tugs soft tissues inward. This is why you see retractions at the lower ribs and breastbone: the more retractions visible, the harder the baby is working. Upper chest and abdominal movements tell a related story. In healthy breathing, the chest and belly rise together. When a baby is struggling, the diaphragm contracts forcefully while the chest wall, which is far more flexible in a newborn than in an older child, collapses inward, producing a rocking see-saw pattern.

Nasal flaring is the automatic widening of the nostrils during inspiration. In preterm infants, activating the muscles around the nostrils reduces nasal resistance by roughly 21 to 23%, lowering the effort needed to pull air through the nose and helping stabilize the upper airway.³ It is a compensatory reflex, not a sign that something has gone wrong with the nose itself.

Grunting may be the most clinically telling sign. A grunting baby partially closes the vocal cords during exhalation, creating back-pressure that keeps the tiny air sacs in the lungs from collapsing at the end of each breath. The baby is essentially generating its own form of positive end-expiratory pressure. When clinicians hear grunting, they know the lungs are having trouble staying open.

How Well the Score Predicts What Happens Next

The practical question for clinicians is whether the number on this scale actually tells you something useful about how sick the baby is and what kind of support they will need. The evidence is a mix of encouraging and humbling.

A prospective cohort study found a statistically significant correlation between the Silverman Anderson score and blood carbon dioxide levels in newborns, though the relationship was modest. More usefully, the score was a strong predictor of whether a baby would need their breathing support escalated within the next 24 hours: about 79% of newborns who scored 5 or higher had their support increased, compared to 28% of those scoring below 5.⁴ That kind of predictive power matters when you are trying to decide whether a baby can stay on a basic ward or needs transfer to a higher level of care.

The score has also been identified as an independent predictor of whether continuous positive airway pressure (CPAP) will fail in very premature infants, meaning the baby will need to be intubated and mechanically ventilated.⁵ Research from South India found that early lung ultrasound scores were significantly higher in newborns who had Silverman Anderson scores of 7 or more, reinforcing that the bedside assessment tracks with objective imaging findings.⁶

Still, the correlation with blood gas values is weaker than clinicians might hope. While higher scores tracked with higher carbon dioxide, the difference in average CO2 between babies scoring above versus below 5 was not statistically significant.⁷ The score is better at flagging who needs more help than at precisely quantifying how abnormal the blood gases are. It functions as a clinical alarm, not a substitute for laboratory analysis.

The Reliability Problem

Here is where the score’s simplicity becomes a double-edged sword. Because each component depends on what a human observer sees and hears at the bedside, different clinicians can arrive at different numbers for the same baby.

A study of preterm infants found poor inter-rater reliability, with an overall agreement measure of just 0.34. Some individual items fared better than others: expiratory grunting showed substantial agreement between raters, which makes sense because grunting is audible and hard to miss. But the chest movement items were much harder to score consistently, with only fair to moderate agreement.¹ Reassuringly, when the same observer scored the same infant on two occasions, consistency was much better (ICC of 0.77), suggesting the issue is not the tool itself but the difficulty of different people interpreting the same physical signs the same way.

A separate study comparing the Silverman Anderson score to the Downes score in preterm newborns found even lower inter-rater agreement for the SA score, describing it as poor (ICC of 0.19).⁸ This is a real limitation. A bedside tool is only as useful as the consistency with which it is applied, and these findings suggest that two nurses looking at the same baby could give meaningfully different scores, especially for the subtler chest-movement items.

Training Closes the Gap

The good news is that structured training substantially improves both accuracy and consistency. A study in a low-resource special care unit found that after a targeted training program, nurses improved their identification of mild respiratory distress from 63% to 93% and of moderate distress from 40% to 73%. Recognition of severe distress was already high before training and held steady. Overall agreement between observers jumped from a kappa of 0.59 to 0.84, mainly because training reduced the tendency to overestimate how distressed a mildly affected baby actually was.⁹

That last detail deserves attention. Before training, nurses tended to err on the side of scoring babies higher than the clinical picture warranted. In a well-resourced hospital, overestimating distress might lead to some extra monitoring. In a low-resource facility where a high score could trigger a risky transfer to a distant referral center, overestimation carries real costs for the baby and the family. Training brought the scores into closer alignment with reality, which translates directly into better decisions.

How the Silverman Anderson Score Compares to the Downes Score

The Silverman Anderson score is not the only bedside respiratory distress scale. The Downes score is probably its most common alternative, and the two overlap but are not identical. The Downes score adds items like respiratory rate and skin color (or, in some versions, oxygen saturation) while retaining similar retractions and grunting items. Because it incorporates additional vital-sign data, it takes longer to complete.

A comparative study of full-term newborns found that the Downes score had somewhat better overall predictive accuracy for respiratory failure (area under the curve of 0.918 versus 0.876) and notably higher specificity, meaning it was better at correctly identifying babies who would not develop failure. The Silverman Anderson score, by contrast, had higher sensitivity, catching a larger proportion of babies who did go on to develop failure. And it was meaningfully faster: about 90 seconds per assessment compared to about 150 seconds for the Downes score.¹⁰

That speed difference matters in an emergency. The same researchers recommended using the Downes score for routine evaluations when time is less pressing, and reserving the Silverman Anderson score for acute situations where a quick read is needed.¹⁰ In practice, many neonatal units pick one tool and stick with it, since switching between scoring systems introduces its own opportunities for confusion and error.

The Score’s Role in Low-Resource Settings

The Silverman Anderson score occupies an interesting niche in global neonatal care. In high-resource neonatal intensive care units, clinicians have continuous pulse oximetry, arterial blood gas monitoring, lung ultrasound, and chest X-rays at their disposal. In those environments, the score is often used as a quick screening complement rather than a primary decision-making tool: a nurse notes a rising score at the bedside and then confirms with objective data before changing the treatment plan.

In low-resource settings, however, the score becomes far more central. Facilities that lack blood gas analyzers or imaging equipment rely on clinical observation to guide care, and the Silverman Anderson score provides a structured framework for doing so. A prospective cohort study concluded that the score may be valuable for predicting the need to escalate respiratory support and for guiding decisions about transferring newborns to better-equipped facilities.⁴ The tool requires no electricity, no consumable supplies, and no calibration. In a district hospital in sub-Saharan Africa or rural South Asia, those characteristics are not minor conveniences; they determine whether any standardized assessment happens at all.

This dual role means the score is likely to remain relevant even as technology advances. In well-equipped hospitals it serves as an early warning system and a common language during shift handoffs. In the rest of the world, it may be the primary means available for deciding whether a struggling newborn needs transfer.

Video-Based AI and Automated Respiratory Monitoring

One of the more intriguing recent developments is the effort to automate parts of the respiratory assessment that currently depend on a human observer. Researchers have developed computer vision systems that use a standard video camera pointed at a baby’s torso to extract respiratory signals. One such system uses a machine learning model to identify the chest and abdominal region, then tracks motion using optical flow analysis to estimate the breathing rate. A pose-based movement detector limits the analysis to periods when the infant is lying still, reducing noise from squirming.¹¹

These systems are not replacing the Silverman Anderson score so much as addressing its central weakness. If the biggest reliability problem is that different observers see different things, a camera watching continuously and measuring motion objectively could help standardize the assessment. The technology is still in early development, and no system has yet been validated as a drop-in replacement for bedside scoring. But the underlying idea, using video to quantify the same chest wall and abdominal movements the score already tracks, is a natural fit. The infants are already being monitored visually; the question is whether a computer can do it more consistently than a tired human at three in the morning.

What the Score Means If Your Baby Is in a Neonatal Unit

If you hear staff mention the Silverman Anderson score while your baby is receiving care, here is what is useful to know. The number runs from 0 to 10. Scores in the low range (0 to 3) generally mean the baby is breathing comfortably or with minimal extra effort. A score in the middle range (4 to 6) signals that the medical team is watching closely and may adjust treatment. Scores at the higher end (7 and above) correspond to significant breathing difficulty, and you can expect the team to be actively intervening.

The score is a snapshot, not a permanent label. Babies can improve rapidly with treatment, and the number can drop within hours once the right support is in place. Serial scoring, where the team reassesses at regular intervals, is common practice and is often more informative than any single measurement. A trend of rising scores is a red flag; a trend of falling scores is encouraging. It is also worth understanding that the score captures effort, not outcome. A baby with a score of 6 who is receiving appropriate breathing support may be doing well and heading in the right direction. The number tells you how hard the baby is working to breathe at that moment, not how the story ends.