Ope n Fract ures of t he

Hand
Review of Pathogenesis and Introduction
of a New Classification System
Jacob E. Tulipan, MD*, Asif M. Ilyas, MD

KEYWORDS
 Open fractures  Hand  Pathogenesis  Classification system  Infection  Treatment

KEY POINTS
 Open fractures of the hand are commonly encountered, and vary widely in mechanism, location,
and severity.
 Current evidence shows that antibiotic use and the extent of contamination are predictive of infec-
tion risk, but time to debridement is not.
 Open fractures of the hand are less susceptible to infection than other open fractures.
 The different regions of the hand are unique with regard to the osseous anatomy, blood supply, and
soft tissue coverage, all of which factor into the risk of infection after an open fracture.
 Current classification schemas for open fractures are insufficient to describe and indicate treatment
of fractures of the hand. A specialized classification is introduced that may better take into account
risk factors for infection specific to the hand when determining best treatment of open fractures of
the hand.

INTRODUCTION become infected, several studies have attempted

to stratify these injuries by infection risk.
Fractures of the finger, hand, and wrist constitute a
significant disease burden, estimated to comprise AVAILABLE EVIDENCE ON OPEN HAND
up to 1.5% of emergency department visits and FRACTURES
constituting 1.4 million cases in 1998 alone.1 Like
all fractures, distal upper extremity fractures range A study by McLain and colleagues2 examined 208
in severity based on several factors, including consecutive patients with open fractures of the
mechanism of injury, fracture location, fracture hand. Overall, the cohort showed an 11% infection
pattern, and associated soft tissue injury. rate. This study had limited subject retention (143
Open fractures of the hand are a common of 208 patients) and excluded both farm injuries
occurrence. A database study in 2001 estimated and human bite wounds. All injuries were irrigated
that 5% of hand fractures are open.1 Like all and debrided in the operating room and received
open fractures, open hand and finger fractures cephalosporin plus/minus penicillin and an amino-
are at increased risk for infection compared with glycoside preoperatively.
their closed counterparts. Beginning with anec- A similar retrospective analysis of factors corre-
dotal observations that these fractures were less lating with infection in open hand fractures
likely than other open fractures of the body to was performed by Swanson and colleagues.3
hand.theclinics.com

This article originally appeared in the January 2016 issue of Orthopedic Clinics, volume 47, issue 1.
Department of Orthopaedic Surgery, Thomas Jefferson University, 925 Chestnut Street, Philadelphia, PA
19107, USA
* Corresponding author. 1025 Walnut Street, Room 516 College, Philadelphia, PA 19107.
E-mail address: [email protected]

Hand Clin 34 (2018) 1–7

https://doi.org/10.1016/j.hcl.2017.09.013
0749-0712/18/Ó 2017 Published by Elsevier Inc.
2 Tulipan & Ilyas

These investigators showed a 6% incidence of The soft tissue coverage of the phalanges con-
infection in a series of 154 patients, with 35 lost sists of skin, tendon, ligament, areolar connective
to follow-up. As in the prior study, all patients tissue, and nail. The 14 phalanges of each hand
were treated with prompt intravenous antibiotics are devoid of muscle. As a result, the digits are
and bedside or operative irrigation and prone to open injury with minimal amounts of
debridement. trauma or fracture displacement, especially in the
An in-depth analysis of functional recovery dorsal surface where the fascial layers lack the
following open fractures in 75 patients performed robustness of the palmar side. Furthermore, these
by Duncan and colleagues4 showed an infection structures do not possess the bulk or vascularity
rate of 6 per 171 fractures (3.5%), all in Gustilo- of muscle, potentially limiting their ability to fight
Anderson type III injuries. This group also under- infection.
went standard treatment with antibiotics and The metacarpals share some morphologic fea-
urgent irrigation and debridement. tures with the phalanges. Among these are palmar
More recent retrospective reviews have varied in layers of tough fascia and alveolar connective tis-
the reported incidence of infection in open hand sue, and a dorsal surface with a thin covering of
fracture. A 2011 review of 145 cases by Capo skin, tendon, and fascia. However, the metacar-
and colleagues5 showed a 1.4% infection rate, pals also benefit from the presence of inteross-
even in a series with a high proportion (91 out of eous, thenar, and hypothenar musculature,
145) of Gustilo-Anderson type III injuries. Similarly, providing bulky coverage and blood supply. As a
a 2006 review of bone grafting for open fractures result, the metacarpals are vulnerable to dorsal
of the hand found a 0% infection rate even in open injuries and wounds but benefit from a robust
more severe fractures.6 Moreover, a 2010 retro- blood supply.
spective review of 432 metacarpal and phalanx The carpal bones possess the most dense soft
fractures requiring internal fixation found no signif- tissue coverage of the osseous regions of the
icant difference in infection rates between the hand. However, they have the most fragile blood
open (133 fractures) and closed (299 fractures) supply because of their absence of muscular
injury groups.7 coverage and otherwise extensive articular nature.
These infection rates are significantly lower Subsequently their blood supply is derived from
than that identified in a 2012 meta-analysis of all their ligamentous and capsular attachments,
open fractures, not only hand open fractures, by structures that can be readily compromised with
Schenker and colleagues.8 That review found an trauma. However, these soft tissue attachments,
8% infection rate in Gustilo-Anderson class I and combined with the deep position of the carpus
II fractures, and a 12.7% rate in class III fractures. and its highly congruent and strong intercarpal at-
This finding supports the traditional wisdom that tachments, provide resistance to open fractures in
the hand is more resilient and less prone to infec- this region.
tion after an open fracture than other open frac-
tures of the body.
Vascular Supply
VARIABLES AFFECTING INFECTION RISK The digits receive most of their blood supply
FOLLOWING AN OPEN FRACTURE OF THE via the palmar digital arteries, with contribution
HAND from the dorsal digital arteries. Distally, these
palmar arteries anastomose to form the blood sup-
There are several potential variables that may ply to the digital pulp.9 The palmar digital arteries
cause an open fracture to be more or less prone run superficial to the digital nerves and lie directly
to developing an infection. These variables include deep to the skin. As a result of their position, these
the local osseous and soft tissue anatomy, the vessels are easily injured during digital trauma,
extent of contamination, the integrity of the soft tis- compromising blood supply and increasing infec-
sue envelope, and the vascularity of the extremity. tion risk of the digit. This effect can be mitigated
by the arterial anastomoses in the digit, which pro-
Anatomy
vide redundant blood supply in case of injury.
Within the hand, distal to the radius and ulna, Degloving, ring avulsion, and other circumferential
there are 27 bones that are prone to injury and injuries are a particular risk for dysvascularity, and
an open fracture. Each has its unique anatomy, loss of both radial and ulnar digital arteries can
blood supply, and soft tissue coverage. Divided result in an avascular digit.
broadly, they can be separated into 3 regions: More proximally, the hand benefits from a
the phalanges, the metacarpals, and the carpal robust and redundant vascularity. The vascular
bones. supply of the hand is provided by the palmar
 Open Fractures of the Hand 3

arches and variable dorsal arches, anastomotic Soft Tissue Envelope

networks composed of contributions from radial
The hand has a unique soft tissue envelope, both
and ulnar circulation. These networks provide mul-
protective and potentially injurious in the setting
tiple perforators supplying both the metacarpal
of open hand fractures. The palmar surface of
bones and the soft tissues surrounding them.
the hand benefits from a robust skin and dense
The intrinsic muscles of the hand also possess
subcutaneous tissue via its glabrous skin and
multiple points of vascular supply, and provide a
deep fascial connections and muscular subcom-
vascular bed that can supply the metacarpals.
partments. Significant trauma is required to result
This region of the hand is more resistant to devas-
in open fractures on the palmar side. In contrast,
cularization from trauma, although extensive
the dorsal hand has only a thin layer of skin with
soft tissue damage can still compromise its blood
minimal alveolar subcutaneous tissue, leaving the
supply.
osseous and tendinous structures prone to ready
Proximal devascularization in hand injuries does
exposure even with minor trauma.
not guarantee loss of blood supply distal to the
When soft tissue loss is present, the hand
injury. Although certain sites in the hand and wrist
poses a unique challenge in coverage. Securing
(eg, the proximal scaphoid) have tenuous vascular
adequate coverage in hand trauma is necessary
supply, the extensive network of anastomoses
to protect the deep osseous and soft tissue struc-
means that blood flow often has many alternate
tures such as the many nerves, vessels, and ten-
paths to reach distal structures. This supply is
dons. However, many soft tissue coverage
especially relevant to the carpus.
options exist that are indicated based on the na-
Because of the extensive articulations of the
ture of the injury and surgeon preference, including
carpal bones, many have a limited, tenuous blood
primary closure, secondary closure, acellular
supply. A landmark study examining 75 cadaver
dermal substitutes, local rotational or advance-
limbs showed that the scaphoid received most of
ment flaps, pedicled flaps, and free flaps. Each op-
its blood supply via distal, dorsal nutrient vessels,
tion has its own unique characteristics and risks
with no intraosseous anastomosis to the palmar
and benefits for infection that must be taken into
circulation. In 70% of capitates examined, most
account in the setting of an open fracture.
of the blood supply was dorsally based and did
not anastomose with the palmar circulation. Like-
Contamination
wise, in 8% of the lunate specimens examined,
the vascular supply of the bone arose from a single Frank contamination of an open fracture intuitively
vessel.10 increases the risk for infection. Contamination of
Although the other carpal bones possess more wounds often occurs as a result of injury mecha-
redundant blood supply, all are vulnerable to nism, because debris is deposited into the wound
disruption from high-energy injuries. In the case site. These contaminant particles provide a nidus
of the scaphoid and the capitate, in particular, for bacterial growth, as well as serving as a source
small soft tissue disruptions may result in avas- of bacterial bioburden. Certain types of contami-
cular bone stock, increasing the risk of infection nation warrant specific consideration. Among
and nonunion. these are soil contamination, which carries a high
Severe vascular injuries of the hand may require risk of anaerobic infection13; fecal contamination,
emergent revascularization, regardless of the which carries a risk of polymicrobial and gram-
level of contamination of the wound. Although the negative infection14; and bite wounds, which may
literature is limited with regard to thrombosis and be contaminated by organisms including Eikenella
infection rate of revascularized hands in open frac- and Pasteurella species.15 A 1978 study per-
tures, inadequate perfusion necessitates emergent formed by Lawrence and colleagues16 analyzed
surgical intervention. Primary repair or grafting of bacterial cultures of open fractures at time of pre-
damaged vessels prevents ischemic injury to distal sentation, and found that infections developed in a
structures, and must be performed if collateral cir- small proportion of patients (3 of 95 fractures), and
culation is not adequate. Poor blood supply is only in those with high levels of contamination,
clearly a risk factor for subsequent infection.11 providing evidence that initial degree of contami-
However, the hand differs to some extent from nation affects infection risk.
other sites of open fracture in its extensive network Tscherne and Oestern17 attempted to quantify
of collateral circulation. Gustilo-Anderson type IIIc this risk with their classification of open fractures.
lower-limb fractures carry an infection risk as high This classification takes into account the severity
as 39% according to one series.12 However, there of associated soft tissue disruption, ranging
are limited data on infection rates in open hand in- from grade I (small puncture wound, negligible
juries with vascular compromise. contamination, low-energy fracture) to grade III
4 Tulipan & Ilyas

(heavy contamination, extensive soft tissue dam- antibiotics. This study did not include information
age, associated neurovascular injury) and grade on the degree of contamination in these injuries.
IV (traumatic amputation).17 A 2015 retrospective The correlation between gross contamination
review of 122 patients by Matos and colleagues18 and infection in the first study discussed earlier
found that Tscherne II and III fractures were asso- provides an indication that this is a significant
ciated with a significantly higher rate of infection contributor to infection risk. Although the distal
(48% and 26% respectively). Although this study radius differs from the hand in soft tissue coverage
examined both upper-limb and lower-limb injuries, and vascularity, this association of injury charac-
it did not differentiate hand injuries specifically. teristics and infection risk can be assumed to be
analogous.
PREDICTIVE FACTORS IN OPEN HAND
FRACTURES INAPPLICABILITY OF THE
A recent meta-analysis of 12 studies on open hand GUSTILO-ANDERSON CLASSIFICATION
fractures meeting the inclusion criteria were TO OPEN HAND FRACTURES
reviewed to assess factors related to infection The Gustilo-Anderson classification11,23 (Table 1),
risk.19 These factors included antibiotic adminis- initially developed for use in long bones, is not
tration and timing of debridement. Use of optimal in classifying hand fractures. Specifically,
antibiotics varied between studies in the meta- the variables used to classify fractures in the
analysis, but all studies using antibiotics used Gustilo-Anderson system, particularly wound
either a cephalosporin or a penicillin derivative. size, and the different nuances of soft tissue
With all patients pooled, antibiotic use was signif- coverage and dysvascularity unique to the hand,
icantly (P 5 .0057) associated with lower risk of make it less applicable to open hand fractures.
infection, with a 4.4% infection rate in the For example, the laceration size cutoffs for
antibiotic-treated group versus a 9.4% rate in the Gustilo-Anderson types (1 cm and 10 cm) are not
control group. Alternatively, timing to debridement realistic for a limb as small as the hand and its fin-
was specifically examined in 2 of the studies used gers. In addition, the indications and options for
in the meta-analysis.2,20 Neither study was able to soft tissue coverage of open long bone fractures
show correlation between timing to debridement (ie, Gustilo-Anderson type IIIB injuries) are very
and infection rate, and nor did the pooled results. different in the hand. Furthermore, there are multi-
Although not specific to the hand, several other ple common mechanisms for open fractures of the
studies have also examined open fractures of the hand. The first is direct laceration or penetrating
distal radius and forearm, and studied different injury. In these cases, a sharp object (eg, a saw)
associated variables relative to infection risk. cuts through skin and then the underlying soft
A 2009 study by Glueck and colleagues14 retro-
spectively reviewed 42 open distal radius fractures
to determine infection risk. Three fractures ulti-
mately became infected, of which 2 were grossly Table 1
The Gustilo-Anderson classification of open
contaminated with fecal matter at the time of
fractures
injury. Although the study found a statistically sig-
nificant correlation between contamination and Type Description
risk for infection, it failed to find any significant as-
sociation between infection and either fixation I Wound <1 cm
method or time to debridement. All 3 infections II Wound >1 cm
occurred in Gustilo-Anderson type II or III injuries. IIIa Extensive soft tissue laceration,
These findings were mirrored in a 2011 study by wound >10 cm, adequate bone
Kurylo and colleagues,21 which retrospectively coverage, segmental fractures
identified 32 open radius fractures. This study IIIb Inadequate soft tissue coverage over
failed to show any infections in the cohort, regard- bone
less of time to debridement or method of fixation. IIIc Arterial injury requiring repair
This study did not report degree of contamination. Data from Gustilo RB, Anderson JT. Prevention of infec-
A 2014 study by Zumsteg and colleagues22 tion in the treatment of one thousand and twenty-five
reviewed 200 open forearm fractures, and found open fractures of long bones: retrospective and prospec-
a 5% infection rate. Deep infection risk was corre- tive analyses. J Bone Joint Surg Am 1976;58(4):453–8;
and Gustilo RB, Mendoza RM, Williams DN. Problems in
lated with injury severity as measured by the the management of type III (severe) open fractures: a
Gustilo-Anderson classification, but was not asso- new classification of type III open fractures. J Trauma
ciated with either time to debridement or time to 1984;24(8):742–6.
 Open Fractures of the Hand 5

Table 2
Proposed classification scheme for open
fractures of the hand and fingers

Type Location Modifiers

I Phalanges a: Primary soft tissue
II Metacarpals coverage not possible
III Carpus (Fig. 1)
b: Frank contamination
(Fig. 2)
c: Dysvascularity requiring
revascularization (Fig. 3)

tissues and bone. In the second mechanism, a

crush injury tears skin while fracturing the bone
below. In the third, shear forces avulse skin and
break underlying bone. In addition, direct blows
or falls can result in a bone spike being forcibly
pushed through the skin. Each of these injuries
can result in similar skin defects (and thus similar
Gustilo-Anderson classes) while causing vastly
different amounts of damage to the soft tissues
and underlying bone.

Fig. 2. A fracture showing frank contamination with

soil. This fracture is classified as type Ib by the pro-
posed classification scheme.

NEW CLASSIFICATION OF OPEN HAND

FRACTURES
Given the current uncertainty with regard to risk
factors for infection and appropriate timing to
debridement of open fractures of the hand, we
recommend the development of a new classifica-
tion system to predict infection risk based on
established risk factors for infection specifically af-
ter an open hand fracture. Subsequently, the clas-
sification system we are proposing (Table 2)
deemphasizes wound size as the primary variable,
and instead takes into account fracture location,
extent of contamination, integrity of the soft tissue
coverage, and viability of the vascularity.
By taking into account both anatomic and injury-
specific factors, this classification can serve as
a more effective tool for guiding treatment by
providing insight for early infection risk stratifica-
tion and long-term prognosis. We recommend us-
ing the classification in the following manner:
Fig. 1. A middle phalanx open fracture showing a
soft tissue defect that cannot be closed primarily.  Any open fracture type (I–III) without a modifier
This fracture is classified as type Ia by the proposed does not require emergent surgical treatment
classification scheme. and can be managed with antibiotics and
6 Tulipan & Ilyas

Fig. 3. An open fracture of the thumb and index metacarpal seen (A) immediately on presentation and (B)
following open reduction and pinning without primary revascularization, showing late necrosis of avascular tis-
sue. Although no direct vascular transection was evident, the traction and degloving nature of the injury resulted
in late vascular compromise and subsequent infection. This fracture is classified as type IIc based on the proposed
classification system and would have potentially benefited from early revascularization.

emergency room washout alone, followed by REFERENCES

standard fracture management.
 Any open fracture type with the modifier “a” 1. Chung KC, Spilson SV. The frequency and epidemi-
requires immediate antibiotics with emer- ology of hand and forearm fractures in the United
gency room washout alone, followed by semi- States. J Hand Surg 2001;26(5):908–15.
elective surgical soft tissue coverage and 2. McLain RF, Steyers C, Stoddard M. Infections in
standard fracture management. open fractures of the hand. J Hand Surg 1991;
 Any open fracture type with the modifier “b” 16(1):108–12.
requires immediate antibiotics and urgent sur- 3. Swanson TV, Szabo RM, Anderson DD. Open hand
gical debridement, followed by early versus fractures: prognosis and classification. J Hand
delayed fracture management. Surg 1991;16(1):101–7.
 Any open fracture type with the modifier 4. Duncan RW, Freeland AE, Jabaley ME, et al. Open
“c” requires immediate antibiotics and emer- hand fractures: an analysis of the recovery of active
gent surgical revascularization and fracture motion and of complications. J Hand Surg 1993;
management. 18(3):387–94.
5. Capo JT, Hall M, Nourbakhsh A, et al. Initial manage-
Using this system and algorithm has resulted ment of open hand fractures in an emergency
in promising and consistent results in our pa- department. Am J Orthop 2011;40(12):E243–8.
tients. It is our expectation that with future 6. Saint-Cyr M, Gupta A. Primary internal fixation and
research this open hand fracture–specific classifi- bone grafting for open fractures of the hand. Hand
cation will be validated as a tool to help guide Clin 2006;22(3):317–27.
early treatment and prognose long-term outcome 7. Bannasch H, Heermann AK, Iblher N, et al. Ten
and infection risk more effectively than the years stable internal fixation of metacarpal and
currently used, nonspecific to hand, open fracture phalangeal hand fractures-risk factor and outcome
classifications. analysis show no increase of complications in the
 Open Fractures of the Hand 7

treatment of open compared with closed fractures. management. J Am Acad Orthop Surg 2015;23(1):
J Trauma 2010;68(3):624–8. 47–57.
8. Schenker ML, Yannascoli S, Baldwin KD, et al. Does 16. Lawrence RM, Hoeprich PD, Huston AC, et al.
timing to operative debridement affect infectious Quantitative microbiology of traumatic orthopedic
complications in open long-bone fractures? A sys- wounds. J Clin Microbiol 1978;8(6):673–5.
tematic review. J Bone Joint Surg Am 2012;94(12): 17. Tscherne H, Oestern HJ. A new classification of soft-
1057–64. tissue damage in open and closed fractures (author’s
9. Park HC, Bahar-Moni AS, Cho SH, et al. Classifica- transl). Unfallheilkunde 1982;85(3):111–5 [in German].
tion of distal fingertip amputation based on the arte- 18. Matos MA, Lima LG, de Oliveira LA. Predisposing
rial system for replantation. J Hand Microsurg 2013; factors for early infection in patients with open frac-
5(1):4–8. tures and proposal for a risk score. J Orthop Trau-
10. Gelberman RH, Gross MS. The vascularity of the matol 2015;16(3):195–201.
wrist. Identification of arterial patterns at risk. Clin 19. Dwyer J, Ilyas A, Ketonis C. Timing of debridement
Orthop Relat Res 1986;(202):40–9. and infection rates in open fractures of the hand:
11. Gustilo RB, Anderson JT. Prevention of infection in a systematic review. Journal of Hand Surgery
the treatment of one thousand and twenty-five 2014;39(9):e44.
open fractures of long bones: retrospective and pro- 20. Ng T, Unadkat J, Bilonick RA, et al. The importance
spective analyses. J Bone Joint Surg Am 1976; of early operative treatment in open fractures of the
58(4):453–8. fingers. Ann Plast Surg 2014;72(4):408–10.
12. Soni A, Tzafetta K, Knight S, et al. Gustilo IIIC frac- 21. Kurylo JC, Axelrad TW, Tornetta P III, et al. Open Frac-
tures in the lower limb: our 15-year experience. tures of the Distal Radius: The Effects of Delayed
J Bone Joint Surg Br 2012;94(5):698–703. Debridement and Immediate Internal Fixation on
13. Templeman DC, Gulli B, Tsukayama DT, et al. Up- Infection Rates and the Need for Secondary Proced-
date on the management of open fractures ures. Journal of Hand Surgery 2011;36(7):1131–4.
of the tibial shaft. Clin Orthop Relat Res 22. Zumsteg JW, Molina CS, Lee DH, et al. Factors influ-
1998;(350):18–25. encing infection rates after open fractures of the
14. Glueck DA, Charoglu CP, Lawton JN. Factors asso- radius and/or ulna. J Hand Surg 2014;39(5):956–61.
ciated with infection following open distal radius 23. Gustilo RB, Mendoza RM, Williams DN. Problems in
fractures. Hand 2009;4(3):330–4. the management of type III (severe) open fractures:
15. Kennedy SA, Stoll LE, Lauder AS. Human and other a new classification of type III open fractures.
mammalian bite injuries of the hand: evaluation and J Trauma 1984;24(8):742–6.