person, whereas no such experience exists for another individual, resulting in no such preference. In this regard, many plastic surgery interventions are sensitive to preference variation (cf., assisted reproduction [43]). Therefore, the combination of having suffering and a preference for plastic surgery is a necessary criterion for publicly funded plastic surgery for non-functional conditions. We mean that suffering needs to be experienced in the sense that a person not experiencing suffering is not suffering [44]. However, suffering can be unwarranted, and we accept that suffering from different causes has different phenomenological qualities. For example, there could be a specific phenomenological quality to suffering related to appearance [45, 46]. A systematic review of physical and psychosocial outcomes after aesthetic surgery indicated that most patients undergoing aesthetic surgery have a lower preoperative QoL (or greater suffering in our terminology) relative to controls, and that the procedure leads to a significant improvement that plateaus over time [34]. Similarly, another systematic review specifically focused on aesthetic rhinoplasty showed that studies suggest an improvement in QoL after the operation [35]. Nevertheless, these studies had methodological shortcomings, such as measurement imprecisions, variability in procedures and outcomes, and heterogenic and small study populations. According to the strong emphasis on person-centered care, in which care should be adapted to the values, experience, and situation of the patient [17, 47], it could be argued that suffering caused by his/her appearance should warrant healthcare (i.e. CJ3). This provides prima facie support allowing the suffering to have sufficient weight in determining whether healthcare should be provided. However, CJ3 is problematic from the perspective of PFE since different patients are differently equipped to report their suffering in a vivid and engaging way, thereby risking inequality. Moreover, it seems CJ3 is also problematic from the perspective of PSE. Even accepting that a problem exists, there might be better treatments for the experienced problem than surgery (e.g., psychotherapy, pharmacological treatment [38], or other psychosocial interventions [48]). Plastic surgery is generally not offered in the public healthcare system to address depression and psychosocial conditions due to bullying, although such patients assign their problems to an appearance feature, and studies show that many people who seek private plastic surgery as adults do so because they were bullied [49–51]. In assessing the size of the healthcare need, we need to assess the size of the gap between the current situation of the patient and the ideal situation (i.e., the problem of assessing the degree of suffering) in a systematic and consistent way to align with both PSE and PFE. Examples of trusting self-reported experience to provide treatment include physical pain, which likely represents the best option for accepting self-reported experience at face value. However, even in these cases, providing a more potent treatment (e.g., morphine) with potential adverse side effects normally requires validation of the pain (e.g., in form of a reasonable biological rationale, such as suffering from cancer). PFE would then require us to do likewise when it comes to nonfunctional conditions and plastic surgery. Sandman and Hansson BMC Medical Ethics (2020) 21:94 Page 6 of 14 Conclusions concerning non-functional appearance conditions (at this point) At this point in the analysis, we draw the following conclusions concerning conditions of a non-functional character requiring plastic surgery [i.e., conditions related to appearance or symptomatic conditions (not affecting function)]. To warrant public funding, the patient needs some form of suffering (and not only a preference for plastic surgery) that must be validated in some form by the healthcare system. This means that the healthcare system must have a reason to believe that the condition results in suffering. Ceteris paribus, the greater the suffering, the stronger the claim for public funding if the suffering is validated. However, at this point, what should guide the validation process remains an open question, given that a clear biological rationale might be lacking (cf., other somatic health conditions). In effect, neither CJ2 or CJ3 are in equilibrium with our three principles PEN, PSE, and PFE. Hence, we need to explore whether CJ4, where the suffering is validated in some form will fare better. Normality as validating non-functional conditions We argued that for functional conditions, statistical normality is a valid criterion for determining whether a condition qualifies for public funding. Following this, we have a prima facie reason from PFE and will move directly from CJ4 to CJ5 (where the suffering is validated by reference to statistical normality) to explore whether CJ5 better aligns with our principles. In order to do so, we first need to distinguish between different uses of normality, where some are ethically unsuitable to be used as guides for reimbursement decisions, i.e. normative normality and normality as etiology. This will leave us with statistical normality as a potential guide also in relation to non-functional conditions. Normative normality? Normality is a complex, value-laden, and, to some extent, historically burdened