Background

Much has been written about the “right” procedure for the repair of pectus excavatum since the introduction of the Nuss procedure [1]. The main focus had been its complication rate in the analysis of the National Surgical Quality Improvement Program for both children [2] and adults [3], also in comparison with Ravitch repair [4]. Although a considerable prevalence of pectus excavatum has been reported in population-based analyses [5, 6], similar data on the epidemiology of funnel chest repair is missing, particularly in adults. We aimed to address this issue by analysing the national hospital statistics of Germany with respect to pectus excavatum repairs in all age groups.

Methods

We obtained aggregated datasets from the Statistisches Bundesamt (Federal Statistics Office) including procedures of the German Modification of the International Classification of Diseases – version 10 for the years 2010 to 2018. The starting year was enforced by the data, because the Nuss procedure was not separately assessed before. We included all codes relevant for the repair of pectus excavatum: conventional repairs (OPS 5-346.a0), implantation of a subcutaneous prosthesis (OPS 5-346.a1), and Nuss procedures (OPS 5-346.a6). Population-based rates of repair were calculated by division of surgical procedures by the official population number on the reporting day of the Statistisches Bundesamt and provided in cases per 100,000 people. This was done separately for the age groups provided within the data. Detailed properties and pitfalls of these data have been discussed elsewhere [7]. The administrative database is derived from hospital reimbursement statistics and covers cases, but not individual patients. Studies using administrative data are exempt from ethical approval, because the case-based data cannot be traced back to the individual patient [8].

Statistical analyses were conducted using R (RRID: SCR_001905) (version 3.5.3) with its generic stats4 package [9], if not stated otherwise. Changes over time were analysed by ordinary least squares linear regression [10,11,12,13], whose requirements of normality of residuals were checked by using the Kolmogorov-Smirnov test and the presence of homoscedasticity was verified by the Breusch-Pagan test, both from the olsrr package (version 0.5.3) [14], aided by visual analysis of QQ plots [15].

Results

Since 2011, the number of conventional funnel chest repairs declined sharply in both males (Fig. 1A, C) and females (Fig. 1B, D), whereas the number of Nuss repairs rose in a similar fashion (Fig. 1F, G). In general, the operations of funnel chest repair with implantation of a subcutaneous prosthesis were rare in patients of both sexes (Fig. 1C, D). The total number of procedures did not change over time, but with an increasing fraction of minimally invasive procedures, the number of males operated on between 15 and 19 years of age rose by 5.3 procedures (F(1,6) = 9.364, P = 0.0222) per year in total. This was however just a shift towards an earlier operation during live, because the procedures in males between 20 and 24 years of age declined in a concomitant fashion by 5.1 yearly procedures (F(1,6) = 23.33, P = 0.0029).

Fig. 1
figure 1

Number of funnel chest repairs in the German population separated by sex and age. A Conventional repairs in males. B Conventional repairs in females. C Conventional repairs with implantation of a subcutaneous implant in males. D Conventional repairs with implantation of a subcutaneous implant in females. E Nuss procedures in males. F Nuss procedures in females

Full size image

Similar changes could not be observed in other age groups in both males and females. On a population-based level, this development did only alter the number of funnel chest repairs per 100,000 males aged 20 to 24 with − 0.2 (95% confidence interval − 0.3 to − 0.09) per year (F(1,6) = 20.42, P = 0.004) and for those in the age group between 25 and 29 years with − 0.06 (95% confidence interval − 0.11 to − 0.02) yearly procedures (F(1,6) = 11.47, P = 0.0147), whereas there were no changes in other age groups in males or in females at all (Fig. 2).

Fig. 2
figure 2

Rates of funnel chest repairs per 100,000 persons. A Rate of minimally invasive and conventional funnel chest repairs per year in males. B Rate of minimally invasive and conventional funnel chest repairs per year in females. Only age groups with more than 0.5 procedures per 100,000 persons per year were depicted

Full size image

Among males in the age groups of 15 to 19 years, the number of Nuss’ procedures increased by 8.7 (95% confidence interval 2.5 to 14.9) per year (F(1,6) = 11.69, P = 0.0142) but did not change in the other age groups (Fig. 3A). This increase was accompanied by a concomitant decrease in conventional repairs in the age group between 10 and 14 years, in which the number of procedures decreased by 3.3 (95% confidence interval 4.9 to 1.7) per year (F(1,6) = 24.59, P = 0.0026) (Fig. 3B). Similar decreases could be found in those aged 15 to 19 years with 3.4 (95% confidence interval 5.7 to 1.1) yearly procedures (F(1,6) = 13.34, P = 0.0107), in those of age 20 to 24 with a similar decrease of 3.4 (95% confidence interval 5.4 to 1.4) procedures per year (F(1,6) = 17.14, P = 0.0061), and patients aged 25 to 29 years, in whom the number of conventional procedures decreased by 1.9 (95% confidence interval 3 to 0.8) per year (F(1,6) = 16.91, P = 0.0063) (Fig. 3B). For the remaining age groups in males, there were no differences in yearly procedures (Fig. 3B). In females, there has been no similar steep increase in the numbers of Nuss’ procedures in all age groups (Fig. 3C), but the data would be more compatible with a slight increase in the age groups between 15 to 19 and 20 to 24, although there is too much variability between the individual data points to clearly support this impression. Decreasing numbers of conventional procedures in females aged 10 to 14 years with 0.8 (95% confidence interval 1.5 to 0.1) per year (F(1,6) = 8.4, P = 0.0274) yearly procedures and a decrease of 0.4 (95% confidence interval 0.8 to 0.008) conventional procedures in females aged 25 to 29 years (F(1,6) = 6.237, P = 0.0467) also provide additional evidence for the aforementioned notion.

Fig. 3
figure 3

Changes in the yearly numbers of minimally invasive and conventional funnel chest repairs in males and females. A Nuss procedures in males. B Conventional procedures in males. C Nuss procedures in females. D Conventional procedures in females. In A and B, only age groups with more than 5 procedures per year were depicted. In C and D, only age groups with more than 2 procedures per year were depicted

Full size image

The substantial increase in minimally invasive funnel chest repairs could also be visualised by the number of Nuss procedures per conventional repair: They increased by 4.2 (95% confidence interval 0.6 to 7.7) per year in 10 to 14-year-old males (F(1,6) = 8.54, P = 0.0266), by 0.5 (95% confidence interval 0.08 to 0.97) in those aged 15 to 19 years (F(1,6) = 8.422, P = 0.0273), and by 0.5 (95% confidence interval 0.09 to 0.9) in males aged 20 to 24 years (F(1,6) = 9.155, P = 0.0232) but did not change in the other age groups in males (Fig. 4A). In females, this was only the case for those aged 25 to 29 years with a yearly increase of 1 (95% confidence interval 0.3 to 1.8) Nuss procedure per conventional repair (F(1,6) = 11.82, P = 0.0138). The data for the other age groups were also more compatible with an increasing relationship between minimally invasive and conventional funnel chest repairs but were too variable to model this increase by ordinary least square regression (Fig. 4B). This was heavily influenced by the small numbers of procedures in females (Fig. 1B, D, E), which made comparative modelling more difficult: For example, there were 18 Nuss procedures in 2018, but no conventional repair, which made the calculation of a ratio mathematically impossible, but highlights the wide adoption of the Nuss repair, although it could not be modelled by linear regression.

Fig. 4
figure 4

Ratio of minimally invasive and conventional funnel chest repairs. A Procedures conducted in males. B Procedures conducted in females. Only age groups in which both procedures were reported were included in the analysis

Full size image

Discussion

Despite a considerable prevalence of pectus excavatum in population-based analyses [5, 6], similarly, also population-based data on the epidemiology of funnel chest repair is missing. We aimed to fill this gap by analysis of the administrative, case-based German national hospital statistics, which has been described to be particularly useful to address such questions [15,16,17]. Using these data, we were able to describe a shift of procedures towards minimally invasive repairs, particularly in males, with a concomitant decrease in conventional procedures, but without an increase in overall procedures. Contrary to other procedures that have been investigated using this dataset [11, 13], in the present analysis, we are able to exclude that procedures have been missed due to their conduction on outpatients or even office-based, as this is impossible for funnel chest repairs.

Although the Nuss repair became much more frequent than conventional repairs in children aged 10 to 14 years, their numbers decreased and contributed towards the substantial increase in procedures in those aged 15 to 19 years. Although Nuss himself first developed his procedure in a 4-year-old boy [1], the median age at his centre has shifted from toddlers, as it was common with the open repair, to 14 years [18]. While its inventors claimed that the minimally invasive procedure might be conducted at any age [1], supported by some centres [19], many others described technical modifications to be frequently necessary for the minimally invasive operation to be successful in adults [20,21,22], including surgeons that operate on children and adolescents, too [23]. The marginal dissemination of such technical modifications might be a reason, why conventional repairs were still prevalent in Germany, particularly in adults aged 25 and older. On the other hand, techniques for open repair of pectus excavatum have enjoyed much popularity in Germany [24] and been advocated by influential surgeons of their periods such as Rehbein’s method [25], the Willital-Hegemann procedure [26], and the Erlangen method [27]. We may thus not exclude that these factors played a role in the dissemination of the minimally invasive funnel chest repair.

Conclusions

We were able to demonstrate a shift towards minimally invasive repair of pectus excavatum in the last decade in Germany, although the total number of procedures did not change in the whole population. Nevertheless, population-based data on the epidemiology of funnel chest repair is missing and we hope to stimulate further research on this matter, particularly from other healthcare systems.