In this cohort study based on a nationwide register, we analysed the trends for surgical treatment of proximal humeral fractures in the entire adult Finnish population. The main finding was that the incidence of surgical treatment of proximal humeral fractures nearly quadrupled between 1987 and 2009. This is of interest as proximal humeral fracture is the third most common osteoporotic fracture type and as such poses considerable strain on our healthcare system. At the same time the incidence of hospitalization due to proximal humeral fractures only doubled, and more specifically, in the oldest age groups the age-adjusted incidence of these fractures has stayed quite constant since the late 1990s
A majority of proximal humeral fractures occur in women with incidence increasing almost exponentially with aging
[19, 20]. According to our study the incidence for surgical treatment rose for both men and women but it is unclear why the rise in incidence is steeper with women. Aging women have shown to have a greater risk than men for an osteoporotic fracture such as proximal humeral fractures
Surprisingly little is known regarding whether two, three, or four part humeral fractures in elderly patients should be treated operatively or conservatively
[8, 11]. There are few randomized controlled trials comparing nonsurgical versus surgical treatment with adequate scoring in follow-up reports
[12–14]. In light of the scarce evidence, the significant increase in plating that occurred after the introduction of locking plates in Finland in 2002 is noteworthy. The number and incidences of ORIF with plating more than doubled between 1998 and 2009. These findings may imply that orthopaedic surgeons adopt new fixation systems without conclusive evidence or knowledge whether these fractures should be treated surgically at all. In a previous independent study we observed a significant increase in the surgical treatment of humeral shaft fractures
. The change in the rate of surgical treatment was not as drastic as in the current study on proximal humeral fractures.
The small number of arthroplasty in the surgical treatment of proximal humeral fractures was surprising as based on the literature, joint replacement is usually suggested especially in age groups of 70 years and older
. The incidence of arthroplasty was quite steady from the late 80’s until the late 90’s. The incidence has since risen (Figure
4) but not as sharply as plating. At the same time fracture plating in women over 70 has gained popularity (Figure
In Finland, medical treatment is equally available to everyone and the study population comprised the entire Finnish adult population; therefore, we consider our study reliable. In addition, previous studies reported the coverage and accuracy of the NHDR injury codes to be over 90%
. A strength of our study is the excellent national coverage of surgically treated proximal humeral fractures; all surgically treated proximal humeral fractures between 1987 and 2009 are included in this study, whether treated as outpatients or inpatients.
A weakness of this study is that the precise incidence of all proximal humeral fractures cannot be assessed using the NHDR data alone because an unknown number of the fractures were treated conservatively on an outpatient basis. Thus we are not able to deduct whether a part of the increase in the incidence of operative treatment of proximal humeral fractures is due to growing numbers of proximal humeral fractures or a growing tendency towards surgical treatment. The available scientific literature suggests that the majority of proximal humeral fractures are still treated nonsurgically
[10, 25]. Another limitation of our 23-year study is the change in the ICD procedure-coding system in 1998. Due to the less specific procedural codes in the ICD-9 system, specific data about the implants (i.e., pinning, plates) used could not be evaluated during 1987–1997. Because of this, the main finding of this study between 1987 and 1997 is the increase in the incidence of surgical treatment of proximal humeral fractures. The implementation of locking plates in Finland occurred at the beginning of the 2000s when the more specific ICD-10 coding system was already in use.
In Finland the use of procedural coding of humeral fracture surgery is exercised as explained in Methods but the practical use of procedural coding between different countries may vary. For instance plating of humeral fracture in Finland is NBJ62 but NBJ61 in Norway. The possible differences in procedural coding have to therefore be taken into account when comparing results between different countries.
According to Bell and co-workers, the incidence of surgical treatment for proximal humeral fractures has increased in North America
. With the lack of consensus on the treatment of choice for proximal humeral fractures, this increased incidence of surgical treatment seems controversial, especially for the older age groups. The lack of evidence makes it difficult to determine whether ORIF with plating is the best surgical treatment option. According to our data, with the exception of plating and arthroplasty, the incidence of all other surgical treatment options has decreased with time, consistent with the findings of Bell et al.