of variation. Following the reasoning in Beauchamp and Childress [15] on the nonmaleficence principle, the harm caused by the breastconserving surgery is a harm justified by the benefits from breast cancer treatment. From a priority setting perspective, this example could be described in a different manner: Woman A suffering from breast cancer experiences problematic symptoms and a high risk of premature death (a highly severe condition). The best and possibly curative treatment for woman A is breastconserving surgery (together with other oncological treatments) along with the possible side effect of a breast malformation. With the breast-conserving surgery, woman A is brought from a very severe condition to a fairly mild one (i.e., the suffering resulting from breast malformation). Another woman, B, born with a discreet malformation also has a slightly more severe condition (a mild-tomoderate condition). Described this way, it seems strange to claim that the healthcare system would have stronger reasons to spend more resources on woman A to improve her situation after having already applied resources for breastconserving surgery than it would to spend resources on woman B with a somewhat worse condition. Hence, normality as etiology risks being in conflict with PSE. It is difficult to identify convincing reasons why accounting for etiology would be ethically relevant when distinguishing how to publicly fund plastic surgery, unless we consider iatrogenic harm resulting from negligence. This does not imply that the described conditions do not warrant public funding, but rather that the etiology of these conditions does not provide an argument for this or apply a special standing to these conditions before other similar conditions with a different etiology. This since it does not align with PSE and PFE. A more promising argument would be that the conditions deviate from what is considered statistically normal. Statistical normality We then focus on statistical normality rather than collectively assumed conceptions of normality, societal normativity, or the etiology of the condition being more or less normal. Statistical normality should be grounded in research and not simply prejudice in order to align with PFE, since we generally require that healthcare is evidence-based. It is often presumed that esthetic conditions, such as facial disfigurements, are worse for females than for males [62], and that females are generally more distressed regarding their appearance [63]. However, recent studies reveal that many men experience body dissatisfaction [64, 65] and request a greater amount of esthetic surgery than previously [66]. Nevertheless, the myth that women suffer more appears to affect practice, as female CLP patients are offered significantly more surgical corrections than male CLP patients [67]. Another example of prejudice is that some conditions are considered more distressing for children than for adults [3, 68]. In practice, some procedures, such as otoplasty, are reimbursed only for children and adolescents and not for adults [3, 68], even though the procedure might reduce suffering in all age groups [69, 70]. There are great variations in specific or sets of physical features in a population. Moreover, some variations are more common than others in both appearance and functionality. From a principled perspective and accounting for a needs-based and publicly funded healthcare system with limited resources, it is reasonable to argue that the greater the deviation in the health of a patient relative to the majority of the population, the greater his/her healthcare need. If a patient has a common variation of health, and most patients are able to live with this variation without treatment – it is reasonable that such a Sandman and Hansson BMC Medical Ethics (2020) 21:94 Page 8 of 14 variation results in lower claims to public resources. This is a standard approach within healthcare, in that it is accepted that patients have different blood pressure and cholesterol levels, knee mobility, and general symptoms within a ‘normal’ range of variation, and that such variation does not constitute a healthcare need that warrants treatment. Within the normal variety, there is a range of different risks of developing more serious conditions but also a range of levels of functionality or suffering. If a patient is within or above the normal range, they represent a positive outlier and do not qualify as having a healthcare need. This hence aligns with both PSE and PFE. How does ‘normality’ relate to the concept of ‘healthcare need’ when considering non-functional conditions? Similar to how statistical normality works in relation to a functional condition, it is also applicable for nonfunctional conditions. Normal variations could be ways to distinguish between non-functional conditions that warrant public funding and conditions that do not, despite the weak relationship between a specific physical feature and suffering, and the fact that a normal range is more difficult to establish. As an example, all female breasts have a certain asymmetry. It is statistically rare within a perfectly symmetrical pair of breasts. Considering greater asymmetries (e.g., those due to underlying malformations, such as Poland’s syndrome [71]), these become more uncommon in the population. Given the need to prioritize scarce