Stop Auditing Mothers and Start Auditing Systems.

National breastfeeding statistics are conventionally read as a measure of maternal behaviour, and the dominant policy response has been to educate and encourage individual women. This Perspective argues that breastfeeding statistics should be interpreted primarily as indicators of system performance rather than maternal behaviour. In high-income countries, much early cessation occurs among women who intended to continue breastfeeding but could not obtain timely, effective support. The resulting shortfall therefore reflects constrained choice as much as personal preference. That both the systems framing and the intention-to-cessation indicator already exist in the literature is acknowledged directly: surveillance research has measured the gap, described its risk factors, and mapped its social patterning. Viewed through a quality-and-safety lens, unwanted early cessation has many of the characteristics of a preventable harm. It is common, patterned by identifiable failures of care, and partly avertable with known interventions such as evidence-based maternity practices, proactive post-discharge contact, adequate paid leave, and protection from substitute marketing. The contribution of this Perspective is not a new metric, but a new use for an existing one. I argue that unwanted early cessation should be treated as an accountable quality indicator within the quality-and-safety systems already used for other preventable harms. Rather than remaining a descriptive statistic, it should become a measure for which health services, employers, and policymakers share accountability. The paper also addresses how intention should be measured, considers the risks of increased surveillance, safeguards maternal autonomy, avoids shifting blame onto clinicians, and distinguishes challenges in high-income settings from priorities in lower-income contexts.