Clinical outcomes following intramedullary nailing of humerus fractures: a systematic review

Abstract

Background

The purpose of this study is to systematically review the literature to assess clinical outcomes arising from intramedullary nailing (IMN) of humerus fractures.

Methods

Two independent reviewers performed a literature search in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Studies were included if they reported on patient-reported outcomes and range of motion following the use of IMN for humerus fractures between 2000 and 2023.

Results

A total of 88 studies, including 4255 patients, met the inclusion criteria for review. The majority of patients were female, with a mean age of 58.3 ± 12.4 years and a mean follow-up of 21.2 ± 15.2 months. The primary outcome score utilized for postoperative assessment was the Constant-Murley Score (CMS). The mean CMS for all fractures, proximal humerus fractures (PHFs), and humeral shaft fractures were 74.7 ± 12.5, 74.2 ± 12.3, and 77.3 ± 12.1, respectively. The mean CMS in 2-part, 3-part, and 4-part PHFs was 80.8 ± 8.8, 80.4 ± 12.3, and 66.5 ± 11.3, respectively. The primary range of motion measurements measured for postoperative assessment were forward flexion and external rotation. The mean forward flexion for all fractures, PHFs, and humeral shaft fractures was 135° ± 16°, 131° ± 13°, and 143° ± 18°, respectively. The mean external rotation for all fractures, PHFs, and humeral shaft fractures was 46° ± 15°, 43° ± 13°, and 52° ± 16°, respectively.

Conclusion

IMN for humerus fractures provides good functional outcomes in both proximal humerus and humeral shaft fractures. However, functional outcomes in PHFs decreased as the number of displaced fracture parts increased.

Humerus fractures, particularly proximal humerus fractures (PHFs), are a major cause of morbidity in elderly patients and are the second most common upper extremity fracture, accounting for 10% of all fractures in patients over 65 years of age.^5,6,54,61 Additionally, humeral shaft fractures account for 20% of all humerus fractures.⁹⁰ Common risk factors for humerus fractures include increased age, female gender, and a history of osteoporosis; falls are the primary mechanism of injury.^5,17,61,111 Humerus fractures in the United States are projected to contribute to nearly 500,000 emergency department visits by 2030; of these, 267,000 will be due to PHFs.⁵⁴

The decision to pursue operative intervention over nonoperative management of humerus fractures depends on patient demographics, fracture characteristics, and patient expectations.³ The surgical options include open-reduction internal fixation (ORIF), intramedullary nailing (IMN), hemiarthroplasty, and reverse total shoulder arthroplasty (rTSA).³ Hurley et al found in a meta-analysis of randomized controlled trials that IMN has demonstrated lower infection rates, reduced operative time, and quicker union as compared to ORIF.⁴⁸ In PHFs, IMN is associated with lower intraoperative blood loss, operative duration, postoperative infection rate, healing time, and surgical trauma, as compared to plating.^24,37,95,97

Despite the growing rates of humerus fractures and the use of IMN as a management option, comprehensive systematic reviews comparing its efficacy across different humerus fracture patterns are lacking in the literature. Given the projected increase in humerus fractures, it is imperative to assess contemporary techniques thoroughly.⁸³ The purpose of this study is to systematically review the literature to compare clinical outcomes following IMN of humerus fractures. The hypothesis was that there will be clinically acceptable patient-reported outcomes (PROs) and range of motion (ROM) among patients undergoing IMN for all humerus fracture patterns.

Methods

Study selection

A literature search was conducted by 2 independent reviewers (K.A.-K. and K.L.) using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.⁶⁴ A title and abstract screen of the search results was performed, followed by a full-text review of potentially eligible studies. Discrepancies between reviewers were moderated by a senior author.

Search strategy

The following search terms were used in the PubMed, EMBASE, and the Cochrane databases in December 2023: (humerus or humeral) and (nail or intramedullary). The search was limited to publications since the year 2000.

Eligibility criteria

The inclusion criteria were the following: (1) clinical studies reporting on postoperative outcomes following IMN of humerus fractures, (2) a minimum of 5 patients, (3) published in English in a peer-reviewed journal. The study with the fewest number of patients who met the inclusion criteria for this paper was 16. The exclusion criteria were the following: (1) biomechanical studies, (2) cadaveric studies, (3) review studies, (4) case reports, and (5) abstract-only studies.

Data extraction

All data were collected by 2 independent reviewers (K.A.-K. and K.L.) utilizing a predetermined Microsoft Excel (version 16.80; Microsoft Corp., Redmond, WA, USA) data sheet. Data included study design, patient demographics, PROs, and ROM. The level of evidence was determined using the Oxford Centre for Evidence-Based Medicine Criteria. Additionally, the methodological quality of evidence was calculated based on the modified Coleman methodology score.¹⁸ Studies scoring between 85 and 100 were categorized as excellent quality; scoring between 70 and 84 was categorized as good; scoring between 55 and 69 was categorized as fair; and scoring below 55 was categorized as poor quality. All studies listed as poor were excluded from the study. Fracture types were categorized into humeral shaft and proximal humeral fractures (further classified according to Neer criteria). Primary outcomes included (1) PROs and (2) ROM.

Statistical analysis

Statistical analysis was conducted using Microsoft Excel. Raw data were collected as either categorical or continuous variables. Categorical and continuous variables were transformed into descriptive statistics to summarize patient demographics, outcome scores, and ROM.

Results

Literature search

The initial literature search yielded a total of 1999 studies. Following a title and abstract screening based on inclusion and exclusion criteria, 421 studies remained. After a full-text review, 132 unique studies met the criteria for inclusion in the final review.^1,2,4,7, 8, 9, 10, 11, 12, 13^,15,16,19, 20, 21, 22, 23^,25, 26, 27, 28, 29^,31, 32, 33, 34, 35, 36^,38, 39, 40, 41, 42, 43, 44, 45, 46, 47^,49, 50, 51, 52, 53^,55, 56, 57, 58, 59, 60^,62,63,65, 66, 67^,69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82^,84, 85, 86, 87, 88, 89^,91, 92, 93, 94^,96,98, 99, 100, 101, 102, 103, 104, 105, 106^,108,112 A visual representation is available in Figure 1.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram. MQOE, methodological quality of evidence.

Study characteristics and patient demographics

Among the 88 studies included, there were 4255 patients. The mean methodological quality of evidence was 68.3 ± 11.0. Overall, 5 studies were in the excellent group, 26 studies were in the good group, and 57 studies were in the fair group. The patient cohort was predominantly female (60.4%), with a mean age of 58.3 ± 12.5 years (range 34.8-82.1 years) and a mean follow-up duration of 21.2 ± 15.2 months (range 8.7-72 months). The study characteristics and patient demographics are shown in Table I.

Table I.

Study characteristics and patient demographics.

	Total
Number of studies	88
Number of patients	4255
Female	60.0%
Age	58.3 ± 12.4
LOE (%)
I	13 (14.7)
II	17 (139.3)
III	16 (18.2)
IV	42 (47.7)
MQOE	68.3 ± 11.0
Follow-up (mo)	21.2 ± 15.2

LOE, level of evidence; MQOE, methodological quality of evidence.

Overall patient-reported outcomes

The Constant-Murley Score (CMS) was reported in a total of 72 studies, including 3004 shoulders, with a mean score of 74.7 ± 12.5. The relative constant score was reported in 22 studies, including 950 shoulders, with a mean of 83.5 ± 7.4. The American Shoulder and Elbow Surgeons (ASES) scores were reported in 32 studies with 903 shoulders. The mean ASES score was 79.0 ± 14.7. Postoperative disabilities of the arm, shoulder and hand (DASH) scores was reported by 20 studies involving a total of 844 shoulders, with a mean score of 21.7 ± 15.6. A full breakdown of PROs is shown in Table II.

Table II.

Overall patient-reported outcomes.

Outcome	Number of studies	Number of patients	Mean
CMS	72	3004	74.7 ± 12.5
Relative constant	22	950	83.5 ± 7.4
DASH score	20	844	21.7 ± 15.6
ASES score	32	903	79.0 ± 14.7
Simple shoulder test	7	202	8.0 ± 1.3
Oxford shoulder score	7	159	35.5 ± 10.4
Neers	7	27	84.0 ± 4.6
JOA score	3	113	81.5 ± 2.3
SANE	4	224	80.5 ± 11.5

CMS, Constant-Murley Score; DASH, disabilities of the arm, shoulder and hand; ASES, American Shoulder and Elbow Surgeons; JOA, Japanese Orthopaedic Association; SANE, Single Assessment Numeric Evaluation.

Overall range of motion

The mean forward flexion was reported in 23 studies with 834 shoulders and was 136° ± 16° postoperatively. The mean abduction was reported in 19 studies with 720 shoulders and was 125° ± 19° postoperatively. The mean external rotation was reported in 20 studies with 664 patients and was 45° ± 15° postoperatively. A comprehensive breakdown of ROM is shown in Table III.

Table III.

Overall range of motion.

Outcome	Number of studies	Number of patients	Mean°
Forward flexion	22	804	135 ± 16
External rotation	19	628	46 ± 15
Abduction	17	654	126 ± 19

Proximal humerus fractures

In the cohort of PHFs treated with IMN, the CMS was reported in 1835 shoulders, with a mean score of 73.3 ± 12.2. The relative constant score was reported for 647 shoulders, with a mean score of 83.5 ± 7.3. In 459 shoulders, a mean postoperative DASH score of 25.4 ± 16.9 was reported. The ASES score was reported for 589 shoulders, with a mean score of 80.4 ± 8.0. Functional outcomes of PHFs are shown in Table IV.

Table IV.

Functional outcome by Neer type.

Outcome	All PHFs (N)	2-part PHFs (N)	3-part PHFs (N)	4-part PHFs (N)
Constant-Murley Score	74.2 ± 12.3 (1794)	80.8 ± 8.8 (360)	80.4 ± 12.3 (365)	66.5 ± 11.3 (139)
Relative constant score	83.5 ± 7.2 (642)	90.3 ± 9.0 (82)	83.0 ± 4.0 (46)	67.9 ± 7.3 (38)
DASH score	25.6 ± 16.9 (454)	21.9 ± 11.8 (106)	20.4 ± 3.8 (188)	32.3 ± 0.8 (14)
ASES score	84.1 ± 8.2 (553)	83.9 ± 7.0 (236)	76.8 ± 11.8 (103)	82.3 ± 6.8 (91)
Simple shoulder test	9.0 ± 1.3 (202)	7.3 (13)	9.1 ± 2.1 (72)	7.4 ± 1.2 (37)
Oxford shoulder score	35.5 ± 10.4 (159)	38.7 ± 12.5 (79)	39 (38)	36 (14)
Neers	84.0 ± 4.6 (27)	86.6 (16)	80.2 (11)	NR
JOA score	81.5 ± 2.3 (113)	NR	NR	NR
SANE	93.0 (23)	93.0 (23)	NR	NR

DASH, disabilities of the arm, shoulder and hand; ASES, American Shoulder and Elbow Surgeons; JOA, Japanese Orthopaedic Association; SANE, Single Assessment Numeric Evaluation; PHFs, proximal humerus fractures; N, number of patients; NR, not recorded.

In the PHF cohort, the mean forward flexion for 607 shoulders was 133° ± 15°. Abduction was reported for 399 shoulders, with a mean of 118° ± 17°. External rotation was reported for a total of 373 shoulders with a mean of 41° ± 13°. Postoperative ROM of PHFs is summarized in Table V.

Table V.

Range of motion by Neer type.

Outcome	All PHFs (N)	2-part PHFs (N)	3-part PHFs (N)	4-part PHFs (N)
Forward flexion°	131 ± 13 (571)	145 ± 19 (144)	144 ± 17 (65)	131 ± 12 (144)
External rotation°	43 ± 13 (337)	46 ± 13 (144)	42 ± 4 (42)	47 ± 17 (61)
Abduction°	124 ± 17 (399)	139 (10)	131 ± 17 (46)	128 ± 19 (87)

PHFs, proximal humerus fractures; N, number of patients.

Humeral shaft fractures

In the cohort of humeral shaft fractures treated with IMN, the postoperative CMS was reported in 1003 shoulders, with a mean score of 77.3 ± 12.1. Postoperative DASH was reported in 335 shoulders with a mean score of 16.1 ± 9.8. In 278 shoulders, the mean ASES score was 74.0 ± 23.3.

The mean abduction was 136° ± 20° in 285 shoulders. Forward flexion was reported in 311 shoulders, with a mean of 143° ± 18°. External rotation was reported in 255 shoulders, with a mean of 52° ± 16°. Functional outcomes and ROM of humerus fractures are shown in Table VI.

Table VI.

Humeral shaft fractures.

Outcome	Number of patients	Mean
Constant-Murley Score	1003	77.3 ± 12.1
Relative constant score	253	85.0 ± 8.9
DASH	335	16.1 ± 9.8
ASES score	278	74.0 ± 23.3
UCLA score	133	28.6 ± 3.0
SANE	201	79.1 ± 12.5
Forward flexion°	227	143 ± 18
External rotation°	255	52 ± 16
Abduction°	285	136 ± 20

DASH, disabilities of the arm, shoulder and hand; ASES, American Shoulder and Elbow Surgeons; SANE, Single Assessment Numeric Evaluation; UCLA, University of California, Los Angeles.

Discussion

This study demonstrated that IMN of PHFs and humeral shaft fractures results in good functional outcomes and ROM. However, as the complexity of PHFs increased from 2-part to 4-part fractures, the CMS, relative constant score, and DASH scores declined. Additionally, ROM in forward flexion and abduction decreased with increasing fracture complexity. These findings suggest that fracture complexity in PHFs negatively impacts postoperative functional outcomes following IMN.

This review highlights that as PHFs progressed from 2-part to 4-part fractures, PROs such as CMS, relative constant score, and DASH scores decreased. These trends align with prior literature. For instance, Whiting et al conducted a comparative study on 2-part and 3-part PHFs treated with IMN or ORIF, where they found no significant difference in forward flexion or external rotation for 2-part fractures; in contrast, 3-part fractures demonstrated higher external rotation in the ORIF group.¹⁰⁷ Additionally, their 2-part group had higher external rotation and forward flexion values as compared to the 3-part group. Our study coincides with Whiting et al's findings by showing decreasing forward flexion as fracture complexity increased from 2-part to 4-part fractures. Notably, the Whiting study had a small sample size for 3-part fractures treated with IMN, eight patients compared to 85 in the ORIF group, and did not include a 4-part fracture cohort. Additionally, Wong et al reported that 2-part and 3-part fractures treated with IMN achieved significantly higher CMS scores than 4-part fractures.¹⁰⁹ Similarly, Kloub et al observed decreasing CMS as fracture complexity increased from 3-part to 4-part fractures.⁵⁵ Boudard et al compared 3-part and 4-part fractures treated with either IMN or ORIF and found no significant difference between the groups in CMS, relative constant score, or DASH scores.⁸ However, their study noted that the IMN group had a higher mean age and longer mean follow-up time, which could influence outcomes. However, their results displayed promising results as IMN was shown to be a viable option for treating 3-part and 4-part PHFs due to comparable PROs to fractures treated with ORIF. Particularly understanding how patients with 4-part fractures fare is crucial, as these patients may be candidates for rTSA, which has demonstrated good functional outcomes, albeit in an older age group compared to this study.⁶⁸ Given the evolving trends in rTSA, including increased rTSA for fracture in younger patients, evaluating functional outcomes in rTSA compared to IMN is critical to evaluate the role that IMN will play in the future.^14,30 Our findings in the context of the literature suggest that while 4-part fractures generally result in lower functional outcomes, IMN remains a viable option with outcomes comparable to ORIF in more complex fractures.

In our study, humeral shaft fractures treated with IMN had higher functional outcomes and ROM than PHFs. PROs such as CMS, relative constant score, and DASH scores were generally higher in the humeral shaft cohort. Our findings are consistent with Den Hartog et al, who compared IMN and ORIF in humeral shaft fractures and found similar DASH scores.²⁵ However, in their study, the plating group exhibited higher CMS, forward flexion, and external rotation showing that plating may still result in superior functional outcomes. Benegas et al conducted a prospective study comparing ORIF and IMN in humeral shaft fractures and found no significant difference in the University of California, Los Angeles (UCLA) scores at 1-year follow-up.⁷ Interestingly, the UCLA scores in their study were slightly lower than those reported in our review despite their patient cohort being approximately 10 years younger. This leads us to believe that subjective satisfaction elements of the UCLA score, such as pain and patient expectations, may significantly influence outcomes in the younger population cohort seen in the Benegas study. Additionally, Putti et al found similar ASES scores between IMN and ORIF groups at final follow-up.⁸⁸ Notably, their study was heavily skewed toward male patients, whereas our review had a predominantly female cohort. This demographic difference may account for some of the observed variations in PROs. Overall, the PROs seen in our study are generally promising and comparable to the PROs seen in the literature and in ORIF; these results show the importance of continued research in IMN of humeral shaft fractures, which will be crucial as new designs enter the market.

While this study could not separate PROs and ROM based on nail generation due to a lack of collectible information, this is an important consideration as different generations may result in outcome differences. While the current findings highlight the efficacy of IMN in humerus fractures, it is important to consider future directions in implant design and surgical innovation. Recent advancements, including expandable, coated, and carbon fiber nails, have shown potential to reduce operative time, decrease radiation exposure, and improve healing rates.¹¹⁰ Additionally, rotator cuff-sparing humeral nails may enhance postoperative ROM and minimize shoulder dysfunction, particularly in elderly patients. These innovations warrant further investigation through randomized controlled trials and meta-analyses to optimize treatment strategies for different fracture patterns.

Limitations

This study had limitations and sources of potential bias, including the limitations of the original studies. The majority of articles were retrospective or nonrandomized. Many studies also failed to separate outcomes by fracture type, which led to them being excluded from the final analysis. Variations among surgical techniques for nailing may also affect patient outcomes. This study did not evaluate the complications associated with these different fracture patterns treated by IMN. Another limitation is that implant type was not analyzed in this study. Additionally, many studies did not describe how ROM was measured, leading to variability in reference axes between studies.

Conclusion

IMN for humerus fractures provides good functional outcomes in both proximal humerus and humeral shaft fractures. However, for PHFs, functional outcomes decreased as the number of displaced fracture parts increased.

Disclaimers

Funding: No funding was disclosed by the authors.

Conflicts of interest: Eoghan T. Hurley reports editorial or governing board membership in the Journal of Shoulder and Elbow Surgery, European Society for Surgery of the Shoulder and Elbow, and Arthroscopy. Jay M. Levin reports stock of stock options in Stryker and Zimmer. Oke Anakwenze reports being a paid consultant for Exactech Inc., LIMA, Responsive Arthroscopy, and Smith & Nephew. They also hold stock or stock options in Suture Tech. Yaw D. Boachie-Adjei reports being a presenter or speaker for DJ Orthopaedics. They also hold stock or stock options in K2M. Christopher S. Klifto reports being a paid consultant for Acumed, LLC, Smith & Nephew, Styker, and Restore3d. They also hold stock of stock options in GE Healthcare, Johnson & Johnson, Merck, and Pfizer. The other authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.