In our study, SES did not seem to impact TTD, even if different markers were used to reflect both individual and ecological deprivation. In the literature, the link between SES and TTD is not consistent. At the ecological level, the same observations were observed in adults and children (22–24). At the individual level, SE variables and TTD can be associated. We reported a trend for children with single-parent families, who presented with an increased TTD. Even if such data are not retrieved in the literature for comparison, difficulties of single-parent families are well described and are associated with deprivation and lower educational levels (25). This is notably the case for single mothers, who experience social and economic disadvantages and lower social support (26–28). These inequalities can make it more difficult to obtain a medical appointment, notably if a financial contribution is necessary. In our study, the father’s profession impacted TTD, and a link between parental education and TTD was observed in developing countries (23, 29–31). Indeed, families with high SES usually benefit from comfortable incomes, which facilitate access to private health insurance (31). Therefore, these families benefit from access to care in several ways, with the potential to visit private hospitals, leading to a decreased TTD (31).
In France, all citizens may benefit from equal access to the healthcare system with no financial contribution (32–34). At least one pediatric oncology center is available per geographic area, which reduces the time to access the hospital. This accessibility to hospital, independent of income conditions, may explain the relatively short TTD of 28 days in France in comparison with developing countries from 47 to 70 days (14, 23, 30). However, our results were similar to those reported in the UK: TTD ranged from two and a half days for children with hemopathies to four days for children with bone tumors between our two cohorts (35).
French private medical demography is well known to be unequal, notably in rural areas where GPs are rarer and cannot cover all necessary care requests (36, 37). However, GPs represented the first physicians contacted in the prediagnosis pathway in our study. These inequalities depend on territories and regions’ attractivity and explain our differences in TTCC (37). Additionally, private clinicians can ask for financial contributions, which may increase social disparities. Therefore, these factors can explain the shorter TTD in children first managed in hospitals, where health professionals are permanently available and care is free (financed by French social security). This free hospital health care can explain the shorter TTD observed in children whose fathers were workers, but these results are exploratory and require further explorations to be conclusive.
In adults, there is an impact of SE disparities on TTD, with an increased probability of severe cancer presentation in cases of long delay (38). Advanced cancer stages at diagnosis are associated with poor prognosis, increased mortality, and potential physical sequelae and make SES a real prognostic factor (11, 38–40). In children, there are mixed results about the relationship between TTD and mortality (35, 41–45). A systematic review concluded on a link among SES, TTD, and survival, whereas our study did not show such an association (46). Our results can be explained by a difference in behavior in adults toward their children versus themselves. Indeed, children and their good health are their parents’ responsibility. Parents are concerned about their children's health, and any symptoms can be a source of anxiety and concern, leading to a strong medical FU and adherence(29). In our study, older children presented with an increased TTD, and same results were retrieved in the literature for patients older than 10 years (23, 41, 47). Indeed, parents monitor their young children very carefully, whereas adolescents are reluctant to alert their parents about their problems and symptoms (41). In addition, an aggressive nature of malignancies is more frequent in young children, leading to a quicker onset of symptoms and therefore a shorter TTD (47).
In our study and the literature, children with CNS tumors presented with a long TTD, with delays from 45 to 60 days (24, 31, 48, 49). This delay for these patients can be explained by the slow-growing nature of these tumors, with usually benign symptoms like headaches and vomiting (49–52). Although a longer TTD is not systematically associated with poor survival, it impacts a child’s QOL through potential neuropsychological sequalae (17, 48). In these children, SES and geographical origin did not impact TTD. In contrast, children with aggressive cancers (blood diseases and solid tumors) presented with a shorter TTD. Indeed, they have fast-growing tumors with a quicker progression of symptoms, which explains the relationship between a shorter TTD and advanced cancer stages (41).
In our study, 12 children used alternative medicine. Even if the use of alternative medicine was not a risk factor for delayed diagnosis in our study, it was reported as such by Handayani et al. (23). We would like to support this hypothesis and be aware of the use of such medicine. Indeed, the low number of children in our group limits the validity of our multivariate analysis. Additionally, a significant difference in TTD by the Wilcoxon test was noted between the groups according to the use of alternative medicine.
Our study presents some limitations. Two centers of the LOGAFTER database did not participate in our study and expose us to a selection bias. We excluded children whose symptoms had occurred more than two years prior to avoid recall bias. However, some children with CNS tumors can presented with extremely long TTD, reported up to 7 years (53). The retrospective nature of our study exposes us to missing data, and few analyses were affected by a lack of statistical power and limited the comparison with the literature. Additionally, only the few classic symptoms were collected to assess the clinical variables, limiting the precision of our data.