Proximal Humerus Fractures: When to Leave It Alone, When to Fix It, and When to Replace It

A Practical Framework for Deciding Between Nonoperative Management, ORIF, and Arthroplasty

Proximal humerus fractures are one of the most common injuries I see, particularly in older adults after a fall. Despite how frequently these fractures occur (accounting for roughly 5–6% of all adult fractures and nearly 10% of fractures in patients over 65) they remain among the most debated injuries in orthopedic surgery. The question of whether to operate, and if so, whether to fix the fracture or replace the joint, does not have a single right answer. It depends on the patient, the fracture, and a careful assessment of the risks on both sides of the decision.

This post walks through how I think about that decision, organized around the same three questions I cover in the office and in the operating room: when to leave it alone, when to fix it, and when to replace it.

Why This Decision Matters

Proximal humerus fractures are not benign injuries. Medicare registry data show that patients who sustain one of these fractures have significantly worse long-term survival compared to age- and sex-matched individuals without a fracture. This is a fragility fracture population, and the fracture itself is often a marker of overall health decline.

The incidence also skews heavily female and increases sharply with age, mirroring the epidemiology of osteoporosis. Understanding the patient's bone quality, physiologic reserve, and functional demands is just as important as understanding the fracture pattern itself.

When to Leave It Alone (and How)

The majority of proximal humerus fractures (approximately 80%) are minimally displaced and

can be managed non-operatively with good outcomes. Even among displaced fractures, nonoperative treatment is often the right choice. The challenge is identifying which patients will heal well and which patients will end up frustrated, stiff, and in pain.

Who Is Going to Heal?

Not all fractures that are treated non-operatively will achieve solid union. A large cohort study of over 2,200 patients found an overall nonunion rate of 10.4% after nonoperative treatment. The single most important predictor was the head-shaft angle (HSA):

• HSA >140° → nonunion rate of only 0.8%

• HSA ≤140° → nonunion rate of 12.4%

Two additional independent risk factors compound that risk: increasing head-shaft translation (HST) and active smoking. A smoker with a 90° head-shaft angle and 75% translation has a predicted nonunion rate approaching 90%. These numbers matter when counseling patients about realistic expectations.

Who Is Going to Be Happy With Nonoperative Care?

Healing without surgery and having a good functional outcome are not the same thing. A separate study focused specifically on predicting function after nonoperative treatment of surgical-neck fractures and found two key radiographic variables:

• Varus angulation (neck-shaft angle on AP in internal rotation): for every 10° of varus, Constant score decreases by 1 point and forward flexion decreases by 4°.

• Greater tuberosity displacement (GTI on the scapular-Y view): for every 10% increase in the tuberosity index, Constant score decreases by 3 points, forward flexion by 7°, and external rotation by 5°.

Using a decision tree model, the patients most likely to achieve good function (flexion >120° and ER >40°) were those with GTI <1.15 and NSA >140°. That group represented the most common fracture pattern, comprising 36% of the cohort. On the other end, GTI >1.15 combined with age over 76 or NSA <119° predicted the worst functional outcomes.

The clinical takeaway is that a well-positioned fracture in a reasonably young patient is very likely to do well without surgery. A significantly displaced tuberosity in an older patient with marked varus deformity is unlikely to achieve a satisfying functional result non-operatively.

The PROFHER Trial (and Why I Don't Use It as the Final Word)

The PROFHER trial is often cited as the definitive evidence that surgery does not improve outcomes for displaced proximal humerus fractures. The RCT enrolled 231 patients with a mean age of 66 and found no significant difference in Oxford Shoulder Score between surgical and nonoperative management at two years.

I take that data seriously but I also think it should be applied carefully. The trial had meaningful limitations:

• Fracture severity: True 3- and 4-part fractures were under-represented, limiting applicability to complex patterns.

• Surgeon experience: Case volume per surgeon was very low, which may not reflect what experienced shoulder surgeons can achieve.

• Selection bias: Patients with obvious surgical indications like fracture-dislocations and severe instability were excluded.

• Implant era: The trial was conducted before reverse shoulder arthroplasty became the dominant surgical option for these fractures in the elderly.

PROFHER tells us that routine surgery for displaced surgical-neck fractures does not automatically beat nonoperative care. It does not tell us that surgery is never the right answer.

How I Rehab My Nonoperative Patients

I follow a straightforward three-phase protocol:

• Weeks 0–4 (Protection): Sling immobilization, elbow and hand range of motion exercises, pendulums as tolerated.

• Weeks 5–8 (Motion): Begin passive-to-active-assisted shoulder range of motion. Supervised or home-based — the evidence suggests both approaches yield equivalent outcomes at 3 and 12 months.

• Weeks 9–16 (Strength): Progressive rotator cuff and deltoid strengthening once adequate motion is established.

When to Fix

ORIF with a locking plate is most appropriate in younger patients (typically under 65) with good bone quality, displaced fractures that exceed acceptable alignment thresholds, and sufficient bone stock to hold fixation. The decision to fix is straightforward when the patient is young and the fracture geometry is favorable. The challenge is avoiding fixation failure, which is common and carries real consequences.

Predicting Failure Before Surgery: Plain X-Ray

The deltoid tuberosity index (DTI) is a simple, underused tool for assessing bone quality from a standard AP radiograph. It measures the ratio of the outer cortical diameter to the inner endosteal diameter at the level of the deltoid tuberosity.

A DTI <1.4 correlates with low bone mineral density and significantly increased risk of fixation failure. Before I commit to locking plate fixation in an elderly patient, I look at this ratio. If the cortex looks thin and the endosteum looks wide, I recalibrate the plan.

Predicting Failure Before Surgery: CT

A separate study using preoperative CT measurement of the humeral head thickness found that the overall screw cutout rate after locking plate fixation was approximately 18%. Head thickness was the key predictor:

• <20 mm: Markedly increased risk of screw cutout.

• >25 mm: Near-zero risk of cutout.

This measurement is straightforward on axial CT and takes about 30 seconds to obtain. If I am on the fence about whether a patient's bone can hold a plate, this number may tip the decision.

Avoiding Failure Intra-operatively

Loss of reduction after locking plate fixation occurs in up to 29% of patients within the first postoperative year and it has meaningful clinical implications. Those who lose reduction have a 44% complication rate and a 30% re-operation rate, compared to 13% and 7% in those who maintain alignment.

The three strongest radiographic risk factors for loss of reduction are:

1. Calcar comminution: the single strongest independent predictor.

2. Low humeral head height: less than 5 mm above the greater tuberosity.

3. Increased screw-calcar distance: ≥12 mm from the most inferior locking screw to the calcar cortex.

Separately, a large retrospective study identified two additional technical factors that surgeons can control intraoperatively:

• Proximal screws >45 mm → 5.3× increased odds of screw cutout.

• Lack of medial calcar support → 4.9× increased odds of screw cutout.

• Lateral diaphyseal translation >6 mm → 2.7× increased odds of loss of reduction.

• Varus reduction → 4.3× increased odds of overall complications.

Reduction quality and medial support are non-negotiables of proximal humerus fixation. A varus construct without calcar contact is likely to fail regardless of which plate or how many screws are used.

A Few Technical Pearls

In my practice, I have found the following consistently helpful:

1. Use the deltopectoral approach rather than an anterolateral split — it provides better exposure and access to the calcar.

2. Control fragments through cuff tissue sutures rather than direct bone contact.

3. Apply the correct reduction maneuver for the fracture pattern: lever for a varus 2-part; derotation for a 3-part; elevation for an impacted 4-part.

4. Ensure proper plate position under fluoroscopy before locking any screws.

5. Consider bone graft for defects at the medial calcar when comminution is significant.

When to Replace

Arthroplasty is indicated when fixation is not going to work, either because the bone cannot hold implants, the fracture geometry is too unfavorable, or the humeral head vascularity is at high risk of compromise.

RSA vs. ORIF: What Does the Evidence Show?

The DelPhi trial was a multicenter, single-blinded RCT comparing reverse shoulder arthroplasty (RSA) to locking plate fixation in 124 patients aged 65–85 with displaced AO/OTA B2 and C2 fractures. At five-year follow-up, RSA was superior on Constant score (mean difference +13.4 points, p=0.002). The difference was consistent across both fracture subtypes, though it was more pronounced in the C2 (impacted, markedly displaced) group.

This is clinically important data. In the right patient (older, low bone quality, complex fracture) attempting locking plate fixation is not a neutral decision. It carries a meaningful re-operation risk that RSA largely avoids.

RSA vs. Hemiarthroplasty: The Debate That Isn't

The SHeRPA trial prospectively randomized patients ≥65 with displaced 3- and 4-part fractures to RSA versus hemiarthroplasty. At 12 months, RSA produced significantly better Constant scores (51 vs. 35) and better health-related quality of life. The takeaway is that in 2026 hemiarthroplasty should not be the choice for acute fracture arthroplasty in the elderly. RSA is the standard.

Does Timing Matter?

A consecutive cohort study comparing acute versus delayed RSA for proximal humerus fractures found similar final functional outcomes between the two groups. The meaningful difference was that acute RSA achieved pain-free function sooner. There is no data supporting delay in an appropriate surgical candidate. If the indication is there, it makes sense not to wait.

A Canadian Evidence Summary

A position statement from the Canadian Shoulder and Elbow Society and Canadian Orthopaedic Trauma Society synthesized the current literature and arrived at several important conclusions:

• Operative versus nonoperative: similar function and pain for displaced fractures overall.

• Locking plates for 3- and 4-part fractures: higher re-operation risk than nonoperative care.

• RSA versus nonoperative: similar function, but better pain relief with RSA.

• Clear surgical indications (low-quality evidence): surgical neck displacement >100% of shaft width and greater tuberosity displacement >10 mm.

• RSA is superior to hemiarthroplasty for both function and pain.

The Bottom Line

Proximal humerus fracture management is not one-size-fits-all, and the temptation to default to a single approach (always operate, or always leave it alone) leads to predictable failures in both directions.

The critical inputs are: fracture geometry on plain X-rays, bone quality quantified by X-ray and CT, displacement thresholds for the tuberosity and surgical neck, and an honest assessment of what the patient is likely to achieve with each option. With that information in hand, the decision to leave it, fix it, or replace it usually becomes clear.

Call to book an appointment to see if PRP might be a good choice for your shoulder or elbow condition:

646-665-6784

---

About the Author

C. Lucas Myerson, MD – Orthopedic surgeon specializing in shoulder and elbow surgery.

Disclaimer

This article is for educational purposes only. It is not a substitute for medical advice. Always talk to your doctor before starting or changing treatment.

Sources

• Fraser AN et al. Reverse shoulder arthroplasty is superior to plate fixation for displaced proximal humeral fractures in the elderly — five-year follow-up of the DelPhi randomized controlled trial. J Bone Joint Surg Am. 2024.

• Haws BE et al. Risk factors for loss of reduction following locked plate fixation of proximal humerus fractures in older adults. Injury. 2023.

• Lapner P et al. Position statement: management of proximal humerus fractures. J Shoulder Elbow Surg. 2022.

• Panagopolous et al. Acute versus delayed reverse total shoulder arthroplasty for proximal humeral fractures — a consecutive cohort study. J Shoulder Elbow Surg. 2022.

• Østegaard et al. Physiotherapist-supervised exercises versus unsupervised home-based exercises after nonsurgically treated proximal humerus fracture — a multicenter randomized controlled trial. J Shoulder Elbow Surg. 2024.

• Spross C et al. Deltoid tuberosity index: a simple radiographic tool to assess local bone quality in proximal humerus fractures. Clin Orthop Relat Res. 2015.

• Stern L et al. Preoperative measurement of the thickness of the center of the humeral head predicts screw cutout after locked plating of proximal humeral fractures. J Shoulder Elbow Surg. 2021.

• Goudie EB et al. Prediction of nonunion after nonoperative treatment of a proximal humeral fracture. J Bone Joint Surg Am. 2021.

• Jacxsens et al. Predicting functional outcome after nonoperative treatment of proximal humeral fractures involving the surgical neck. J Shoulder Elbow Surg. 2025.

• Handoll et al. Surgical versus nonsurgical treatment of adults with displaced fractures of the proximal humerus — the PROFHER randomized clinical trial. Lancet. 2015.