1 Introduction

Proximal humerus fractures (PHFs) are the third most common fractures in adults. They occur in a bimodal distribution, with the first peek in a young patient population following high-energy mechanisms, and the second peak in the elderly osteoporotic population from typically low energy mechanisms. Within both these groups are patients who require their upper extremities for mobilization and ambulation, whether due to polytrauma or pre-injury assisted ambulation. These patients have been termed “functional quadrupeds”, a phrase coined by the senior author and previously published.¹

Despite the massive amount of literature written on the topic, treatment of PHFs is an area of much debate and much variability between practitioners.^2,3 Because of this, multiple survey-style studies have looked at treatment attitudes.^4–6

Further, PHFs are unique in that they are frequently treated by two separate subspecialties of orthopedics: orthopedic traumatologists (OT) and shoulder and elbow (SE) surgeons. Prior studies have compared these groups as well. One study found significant differences in treatment preferences between the subspecialties in only two of five geriatric PHF clinical scenarios.⁴ Another found poor intraobserver agreement on treatment choice within both subspecialties, and that shoulder surgeons were statistically more likely to perform arthroplasty.⁷ Multiple studies have shown shoulder and upper extremity surgeons were significantly more likely to choose operative management than OT.^7,8 However, while multiple surveys have evaluated attitudes toward treating PHFs and even compared preferences between shoulder and elbow and trauma surgeons, none of these surveys assessed differences in postoperative protocols amongst the subspecialties, amongst the treatment options, or based on patient characteristics such as concurrent injuries or pre-injury ambulatory status. We felt postoperative protocols are an under-studied aspect of caring for these patients, given the high percentage of functional quadrupeds with PHFs.

We thereby surveyed OT and SE surgeons regarding their treatment patterns for PHFs. The goals of this study are to evaluate standard postoperative protocols following surgical treatment of PHFs with open reduction internal fixation with a locking plate (ORIF), intramedullary nailing (IMN), or arthroplasty, and evaluate which factors orthopedic shoulder and elbow and trauma surgeons consider when deciding how to manage PHFs.

2 Materials and methods

This study was exempted from IRB approval by our institution. A web-based questionnaire was developed based on review of prior survey studies on proximal humerus fractures. It was edited by one OT surgeon and one SE surgeon for wording, relevance, ease of understanding and redundancy. It included demographic and practice structure questions about respondent's subspecialty, region of practice, years in practice, number range of acute proximal humerus fractures treated per year, and modalities for acute proximal humerus treatment used in the past year. Surgeons were asked their standard postoperative protocols for weight bearing, lifting, and range of motion (ROM) following treatment of an acute PHF with IMN, ORIF, and arthroplasty (including reverse total shoulder arthroplasty [rTSA] or hemi-arthroplasty [HA]). Treatment considerations were evaluated by factors affecting chosen treatment modality and factors affecting postoperative protocols.

The survey was administered using Google Survey by providing an internet link. This link was distributed to members of the American Shoulder and Elbow Surgeons (ASES), AO North America Trauma, and Orthopaedic Trauma Association (OTA). Monthly emails were distributed which encouraged members to participate. Email addresses were collected to ensure unique responses. Data was otherwise anonymous. All attending orthopedic surgeons willing to participate were eligible for inclusion.

We analyzed data using SPSS software Version 25.0 (IBM Corporation, Armonk, NY), which was used to generate descriptive statistics for all variables. All respondents irrespective of subspecialty were included for descriptive statistics. For comparative analysis, only OT and SE surgeons were included. We utilized Pearson's Chi Square (χ2) test to compare categorical variables based on subspecialty (i.e., OT vs. SE), years in practice, and geographic region of practice. P-values less than 0.05 were considered statistically significant.

3 Results

3.2

3.2 Postoperative protocols

Standard immediate postoperative protocols for ROM, lifting, and weight bearing were significantly different between OT and SE surgeons following treatment with IMN, ORIF, and shoulder arthroplasty, with OT surgeons more likely to allow early ROM, lifting and weight bearing after all three modalities. Time from surgery until first decrease in postoperative restrictions was significantly different between OT and SE surgeons following treatment with ORIF but not arthroplasty or IMN, with OT surgeons advancing patients earlier on average. Time from surgery until unrestricted use of the operative extremity was significantly different between OT and SE surgeons following treatment with IMN and shoulder arthroplasty but not ORIF, again with OT surgeons tending to permit full use at an earlier time point [Fig. 1]. Separate analyses by years in practice and by region of practice showed no significant difference between postoperative protocols for ROM, lifting, weight bearing, postoperative period to first restriction advancement, or postoperative period to discontinuing all restrictions following treatment with IMN, ORIF, or shoulder arthroplasty (all p > 0.05). An analysis of number of PHFs treated per year showed no significant differences in postoperative protocols for ORIF or arthroplasty (all p > 0.05). For IMN, only ROM was significantly different, without clear pattern based on volume per year (p = 0.017).

Close

Fig. 1 Standard Proximal Humerus Fracture Postoperative Protocols by Treatment Modality and Specialty. Rows: IMN = intramedullary nail fixation; ORIF = open reduction internal fixation with locking plate; Arthroplasty = hemiarthroplasty or reverse total shoulder arthroplasty. Columns: Time to First Advancement = weeks postop when first increase in permitted use of the operative limb occurs; Time to Unrestricted Use = weeks postop when patient is permitted unrestricted use of operative arm. Blue = Shoulder and Elbow surgeons, Orange = Trauma surgeons.

Export to PPT

3.5

3.5 Preferred mode of treatment

Fig. 2 highlights the mode of treatment in which surgeons would feel the most comfortable allowing no postoperative weight bearing, ROM, and lifting restrictions. There was a significant difference between OT and SE surgeons in the treatment modality where they would feel most comfortable allowing immediate post-op weight bearing, ROM, or weight bearing as tolerated, with trauma surgeons preferring IMN and SE surgeons preferring reverse total shoulder arthroplasty. Separate analyses by years in practice and by region of practice showed no significant differences in treatment modalities preferences for increasing weight bearing, ROM, or lifting immediately post-operatively. Analysis by number of PHFs treated per year showed no significant differences in treatment preferences for increasing weight bearing or lifting immediately post-op, but did show significant differences for ROM (p = 0.013) with those treating more PHFs per year more likely to choose rTSA.

Close

Fig. 2 Mode of Treatment After Which Surgeons Would Feel Most Comfortable Allowing No Post-Operative Restrictions When Treating an Acute Proximal Humerus Fracture. IMN = intramedullary nail; hemi = hemiarthroplasty; ORIF = open reduction internal fixation with locking plate; rTSA = reverse total shoulder arthroplasty. Blue = Shoulder and Elbow surgeons, Orange = Trauma surgeons.

Export to PPT

4 Discussion

We found significant differences in postoperative protocols between OT and SE surgeons following treatment of PHFs with IMN, ORIF, and arthroplasty. OT were more likely to perform IMN, more likely to allow lifting, weight bearing, and motion postoperatively for all modalities, more likely to advance patients’ restrictions earlier in the postoperative course, and more likely to prefer IMN fixation if they wanted to minimize postoperative restrictions. SE surgeons were more likely to perform rTSA and to prefer this if they wanted to minimize postoperative restrictions.

Clement et al.⁹ reviewed 637 PHFs, largely low-energy minimally displaced or two-part fractures, in patients older than 65 years. They found that one-year mortality was 10 %, with the only independent predictor of survival being whether a patient lived in their own home prior to injury. Further, poor social independence correlated with poor clinical outcome, measured by Constant score, irrespective of age and fracture severity. Wang et al.¹⁰ assessed 281 patients older than 65 years who were admitted with humeral fractures, with 84.3 % representing PHFs. Of these, 12.8 % underwent operative management. Primary outcomes were mortality and new admissions to nursing facilities. They found in-hospital mortality of 3.6 %. Furthermore, 17.8 % of patients were newly admitted to a nursing facility upon discharge, with age and Charlson Comorbidity Index (CCI) serving as significant predictors. Myeroff et al.¹¹ reviewed 319 patients older than 60 years who presented to the emergency department (ED) with an isolated PHF. The management (i.e., operative versus nonoperative management) and disposition (i.e., admission to the hospital or discharge home from the ED) of these patients varied. Primary outcomes were one- and two-year mortality. At one-year follow-up, CCI, body mass index (BMI), and ASA score were significant predictors of mortality. At two-year follow-up, CCI, BMI, and reliance upon an ambulatory-assistive device at time of fracture were significant predictors of mortality. Of the patients who were admitted to the hospital, 64 % were discharged to a higher-level living facility than their pre-fracture situation. Twenty percent of the patients experienced a functional decline by at least one level of assisted ambulation at one-year post-injury. There was a correlation between the injury's Neer classification and admission but no correlation between Neer classification and mortality or loss of living/ambulatory independence.

These studies suggests that pre-injury functional status, not chronologic age or fracture morphology, is a risk factor for morbidity and mortality in geriatric patients following a PHF. Furthermore, these results suggest that geriatric populations experience decreased function and increased mortality following a PHF, in analogous fashion to after hip fracture. However, there is no prospective data on how treatment modality of a PHF affects mortality, living status or ambulatory status. There are also no prospective data on safety and outcomes following different postoperative protocols for PHFs. These questions are especially important when treating certain subsets of patients, such as polytraumatized patients and functional quadrupeds who need their upper extremities to ambulate.

The results of the current study must be interpreted within its limitations. As with all survey studies, this study is prone to selection bias, with participants with strong opinions regarding proximal humerus fracture treatment more likely to respond. Further, question design can limit respondent's ability to express the nuances of their treatment preferences. For example, one respondent reported different postoperative protocols for rotator cuff equivalent fractures such as greater tuberosity fractures, compared to surgical neck fractures. Our answer options were multiple choice and close-ended by design, which could have limited respondents' ability to describe treatment distinctions. Finally, as OT and SE surgeons practice in different environments and on different patient populations, differences in standard protocols could have been due to differences in typical injury types treated.

5 Conclusion

When treating a PHF, the decision to operate and which operative modality to use remain subject to significant variability between OT and SE surgeons. Our results demonstrate that postoperative protocols and modes of treatment vary significantly between these two subspecialties. Further research is needed to study postoperative protocols after surgical management of a PHF to optimize outcomes and minimize the rate of functional decline, morbidity, and mortality.

Funding/sponsorship

This study was distributed to surgeons via the American Shoulder and Elbow Surgeons (ASES), AO North America Trauma, and Orthopaedic Trauma Association (OTA).

Informed consent

Not applicable.

Instutional ethical committee approval

This study was exempted from IRB approval by our institution.

Authors contributions

JCK: conceptualization, investigation, methodology, project administration, resources, supervision, validation, review/editing final draft. SN: conceptualization, investigation, methodology, project administration, resources, supervision, validation, review/editing final draft. BRF: conceptualization, data curation, formal analysis, investigation, methodology, validation, writing original draft, review/editing final draft. JS: data curation, formal analysis, investigation, methodology, validation, writing original draft, review/editing final draft. REC: data curation, formal analysis, investigation, methodology, validation, writing original draft, review/editing final draft.

Previous Presentation(s)

Presented as a poster at American Shoulder and Elbow Surgeons (ASES) Annual Meeting 2019.

IRB approval

Exempt per institutional criteria.