Hand Surgeons Are Tackling Tougher Scaphoids: A Study of ABOS Candidate Data

Abstract

Background

This study will evaluate whether those who have completed a hand fellowship treat a disproportionate number of scaphoid fractures based on recent American Board of Orthopaedic Surgery (ABOS) data. We hypothesize that surgeons who have completed a fellowship in hand surgery will address most surgically treated scaphoid fractures, particularly those with fracture nonunions or requiring graft.

Methods

We queried the ABOS database for case log information submitted for part II of the ABOS examination. This search included all cases with Current Procedural Terminology codes for procedures related to scaphoid fixation. Demographic information, case volume, type of case, and complication rate were compared for hand fellowship–trained surgeons and those who had completed other fellowships as well as non-fellowship-trained surgeons.

Results

During the study period, 1686 surgeons reported treating 4244 scaphoid fractures. Of these surgeons, 1180 had completed a hand surgery fellowship. Hand fellowship–trained surgeons were shown to have operatively treated more scaphoid fractures both in total volume and on a per-surgeon basis. Hand fellowship–trained surgeons were also found to have performed a significantly higher proportion of difficult cases, which were those listed as being a malunion/nonunion or those incorporating a pedicle graft. There was no difference in the complication rate between the 2 groups.

Conclusion

Among those orthopedic surgeons reporting case information for part II of the ABOS certification examination, statistically significant differences exist in case volume and case difficulty among surgeons with different areas of fellowship training. Complication rates increase with patient age and examination year.

Background

Scaphoid fractures are the most common fractures (70%) of the carpus. ¹ Furthermore, scaphoid fractures are prone to delayed or nonunion due to tenuous blood supply at the proximal pole, high incidence of synovial fluid interposition, and tendency of the fragments to fall into humpback deformity.^2,3

Determination of appropriate operative or nonoperative treatment has historically been based on fracture pattern, displacement, and potential for healing.^2-4 More recently, there has been an increasing trend in both clinical practice and literature to advocate for acute fixation of even nondisplaced fractures in the acute setting. ⁴ This is supported by a decreased time to bony union;^4,5 improvement return to sport, ⁶ work,^4,5 and military duty; ⁷ and decreased cost ⁵ when treating these fractures operatively.^4-7 While surgical treatment may be favorable in terms of functional outcome and return to work, surgical management does result in a higher complication rate as compared with nonoperative management.^4,8-10

This analysis will assess the current state of scaphoid fracture treatment among surgeons who are early in their careers. In addition, this study will further evaluate whether those who have completed a hand fellowship treat a disproportionate number of these fractures with fixation. We hypothesize that young surgeons who have completed a fellowship in hand surgery will address most surgically treated scaphoid fractures, particularly those requiring vascularized pedicles and fracture nonunions as compared with non-fellowship-trained hand surgeons.

Methods

To become board certified by the American Board of Orthopaedic Surgery (ABOS), one must first pass a written examination (part I). Candidates then practice for 22 months and collect cases during a 6-month period. Candidates then become eligible to sit for an oral examination (part II). During part II, the examinee must present cases from their own practice, selected by the ABOS for review. The ABOS maintains data on the case diagnosis (International Classification of Diseases [ICD] 9 or 10), procedure performed (Current Procedural Terminology [CPT]), patient and surgeon demographics, and complications.

The ABOS part II database was queried for scaphoid fractures of the waist or distal pole (ICD-10) and proximal pole (ICD-10) undergoing open reduction internal fixation (25628), with local bone graft (25440) or with a vascular pedicle (25430). Overall case data were then compiled and sorted using both CPT diagnosis codes and ICD-9 and ICD-10 codes. The data were also subdivided into groups based on whether the surgeon reported hand surgery fellowship, a fellowship other than hand surgery, or no fellowship at all. The reported ICD-9 and ICD-10 codes were also grouped according to the level of difficulty of the cases performed. Those codes which relate to acute fracture involving the waist, distal portion, or unspecified portion of the scaphoid were combined and considered the less difficult procedures. Cases with ICD codes related to acute proximal pole fractures or cases of nonunion were considered more difficult. These groupings were then used as a proxy for evaluating which groups of surgeons were doing more of the difficult cases. The determination of more versus less difficult procedures was made by the 2 senior authors, both of whom are practicing, board-certified, and fellowship-trained hand surgeons. These groupings are thought to accurately reflect the opinion of most practicing surgeons, although this cannot be verified by this study.

Demographic data for both patient information and surgeon information were evaluated and compiled. The primary outcome measure were surgeon-reported complications. Diagnoses and procedures were reported between fellowship-trained hand surgeons and other orthopedic surgeons. The number of cases based on the above ICD-9 and ICD-10 groupings was compared between hand fellowship–trained surgeons and all others to determine which group was doing more difficult cases. Reported complications were compared between the 2 groups to evaluate for potential differences in the complication rate.

Statistical Analysis

When reporting complications, related items were grouped such that the pain category included pain and chronic regional pain syndrome; fracture included fractures and malunion; and vascular included vascular, hemorrhage, and limb ischemia. Any complication category that had less than 4 reported occurrences, which included tendon or ligament injuries and compartment syndrome, was grouped under surgical other. Multiple other factors were also analyzed to determine effect on the complication rate. These factors include patient age, sex, region, examination year reported, and time to follow-up. The statistical analysis was performed using Fisher exact testing with the alpha value set at 0.05. When evaluating factors as they relate to complications, multivariable logistic regression was used with surgical complication as the dependent variable.

Results

There were 4244 reported cases performed by 1686 surgeons. Of these surgeons, 1180 (70%) were hand fellowship trained, while 233 performed a fellowship other than hand and 273 were not fellowship trained. On a per-case basis, hand fellowship–trained surgeons performed the majority of these cases (3.0 hand vs 1.3 nonhand). Within the study population, we found that most of the patients were men and within the 18- to 30-year age group. (Table 1) The majority of the cases were performed in the Midwest. When evaluating surgical complications, nonunion was by far the most commonly reported. (Table 2)

Table 1.

Patient Demographics of Study Sample (n = 4244).

Demographic	N	%
Age, y
<18	683	16.1
18-30	2258	53.2
31-50	965	22.7
>50	338	8.0
Sex
Male	3587	84.5
Female	657	15.5
Surgeon training
Hand fellowship	1180	70.0
Other fellowship	233	13.8
No fellowship	273	16.2
Surgeon region
Midwest	330	19.6
Northeast	323	19.2
Southwest	307	18.2
Southeast	290	17.2
South	243	14.4
Northwest	153	9.1
Other	40	2.4

Table 2.

Surgical Complications (n = 600).

	N	%
Nonunion	234	39.0
Surgical unspecified	106	17.7
Implant failure	61	10.2
Stiffness/arthrofibrosis	47	7.8
Infection	44	7.3
Fracture	44	7.3
Nerve palsy	41	6.8
Pain	9	1.5
Vascular injury	7	1.2
Graft-related problem	6	1.0

Hand surgeons performed most routine internal fixation cases and were also more likely to use local bone graft and a vascular pedicle (P < .0001) (Table 3) When evaluating cases by ICD-9 and ICD-10 codes, hand surgeons were more likely to fix routine fractures as well as proximal pole and fracture nonunions (P < .0001) (Table 4)

Table 3.

Number of Cases and Case Type per Surgeon Grouping.

	No. of cases	No. of surgeons			P value
		Total	Hand	Nonhand
All cases	4244	1686	1180	506	<.0001
Cases by Current Procedural Terminology
Scaphoid fixation (25628)	2855	1455	1006	449	<.0001
Scaphoid nonunion fixation with local bone graft (25440)	1200	708	626	82	<.0001
Vascular pedicle in addition to bone fixation (25430)	51	44	43	1	.663

Table 4.

Cases by ICD-9/10 and Type of Surgeon Performed.

	Hand	Nonhand	P value
ICD-9/10 grouping	N (%)
Acute involving waist, unspecified, or distal portion of scaphoid	2136 (80.30)	524 (19.70)	<.0001
Acute proximal pole and all nonunions	1061 (91.47)	99 (8.53)	<.0001

Note. ICD = International Classification of Diseases.

There was no observed difference in the complication rate between hand and other surgeons when comparing all cases. This was also true when cases were subdivided into groupings based on ICD codes (Table 5). The complication rate was found to increase with patient age (P < .0001). Progressive examination year correlated to an increased complication rate (Figure 1). Patient sex (P = .62) and region in which the surgery was performed (P = .52) had no impact on the complication rate.

Table 5.

Comparison of Complication Rates When Grouped by ICD-9/10 Codes.

	Hand	Nonhand	P value
ICD-9/10 grouping	N (%)
Acute involving waist, unspecified, or distal portion of scaphoid
Surgical complication	274 (12.8)	56 (10.7)	.21
No surgical complication	1862 (87.2)	468 (89.3)
Acute proximal pole and all nonunions
Surgical complication	184 (17.3)	16 (16)	.89
No surgical complication	877 (82.7)	83 (83)

Note. ICD = International Classification of Diseases.

Figure 1.

Linear regression of complication rate by examination year with the trend toward the increasing rate of complication with progressive examination year.

Discussion

The purpose of this investigation was to evaluate trends within scaphoid fracture fixation among surgeons early in their careers. There were 2 key findings in this analysis. First, hand surgeons treated most scaphoid waist (80%) and proximal pole fractures (91.5%) and performed many of the bone grafting procedures. Second, despite treating more complex fractures with more technically demanding procedures, the complication rate was the same between hand surgeons and other surgeons. Given the increased volume and propensity to address more difficult cases while maintaining an equivalent complication rate, it would appear that scaphoid fracture fixation is best performed by a fellowship-trained hand surgeon. Multiple prior studies have demonstrated a correlation between increased volume and improved outcomes pertaining to orthopedic surgical procedures, and while one cannot make a direct comparison between these procedures and scaphoid fixation, it stands to reason that increased exposure would portend better outcomes.^11,12

The present analysis found that hand surgeons fixed most standard scaphoid fractures (69.1%), those with local bone graft (88.4%), those with a vascular pedicle (97.7%), and proximal pole fractures (91.5%). Hand surgeons are more likely to opt for internal fixation of scaphoid fractures. In a recent survey of 494 orthopedic surgeons on their preference in treatment of specific nondisplaced scaphoid fractures, younger surgeons (<45 years old) and hand specialists were more likely to advocate for internal fixation. ¹³ Similarly, hand surgeons are more likely than other orthopedic surgeons to choose internal fixation of distal radius fractures. ¹⁴ In addition, a review of the ABOS case logs for distal radius fracture treatment found that hand fellowship–trained surgeons treated a higher percentage of complex intra-articular fractures.^15-17 Specialization within a hand fellowship and frequent practice of small bone fracture fixation make hand surgeons uniquely capable of operative treatment of these fractures.

Despite tackling tougher scaphoid fractures, hand surgeons have an equivalent complication rate to other orthopedic surgeons. In the aforementioned ABOS case log analysis of distal radius fracture fixation, while hand surgeons operated on more complex distal radius fractures, hand fellowship–trained surgeons had a higher complication rate (12.5% vs 10.7%, P < .05). ¹⁵ However, hand surgeons were more likely to perform concomitant procedures like carpal tunnel release, which may further indicate the unique complexity of the cases performed by hand surgeons while introducing additional sources of complications. ¹⁵

This analysis of young surgeon’s distal radius fracture fixation did not evaluate a trend toward more complications later in the reporting years. ¹⁷ In the present analysis, there was a trend in increasing complication rate with progressive examination year indicating that young surgeons reported an increasingly higher complication rate with each year. This is similar to trends seen in previous ABOS studies. Although it cannot be proven, it is thought that this shift is due to an increase in the number of complicated procedures being performed.^18,19 An analysis of orthopedic surgeons performing shoulder arthroscopy and sitting for part II of the ABOS determined that those who performed a fellowship reported higher complications than those who did not. ¹⁹ It is possible that surgeons being reviewed by a subspecialty ABOS board are acutely concerned about underreporting complications in front of an expert panel. Also, specialized surgeons who best understand the case may be more likely to identify and understand complications when they occur.^20,21

One of the main limitations in this study is the ambiguity in reporting of complications. It is possible that some medical complications are underreported, whereas some surgical complications are overreported. In addition, the data set is not yet able to project outcome measures. However, starting in 2019, prospective data were collected from patients by the ABOS during the time of the case collection. It is possible in the future that these data will be available to study. Also, it was not possible to compare operative and nonoperative management. The groupings of more and less difficult cases by ICD-9 and ICD-10 codes were made based on author opinion and could therefore be seen as a weakness. However, as hand surgeons performed more of the procedures in both groups, the impact of this limitation seems to be insignificant. This grouping also attempts to overcome limitations acknowledged in previous ABOS, namely, the inability to account for case complexity in ABOS data sets. ¹⁹ Despite these limitations, the use of the ABOS data set allows for a snapshot into the trends of practice of young surgeons in the United States. The present analysis determined that hand surgeons fix the vast majority of scaphoid fractures, particularly the more complex proximal pole fractures, and are likely to be best suited for treatment of these complex injuries.