Nonoperative Management of Hand Fractures

3 Nonoperative Management of Hand Fractures

Grey Giddins

Abstract

Historically most hand fractures were treated nonoperatively. While operative techniques have improved considerably, most fractures still do so well with nonoperative treatment that surgery is not warranted. There are strongly conflicting opinions in the published literature on the treatment of some hand fractures. This review aims to reconcile those and highlight subset bias, which may be skewing opinions.

A literature review was performed; spiral or long oblique metacarpal fractures and metacarpal neck fractures treated with mobilization, and transverse metacarpal shaft fracture treated with initial splintage or plaster support and then mobilization all do so well that currently surgery cannot reliably improve the outcomes, so the risks of surgery outweigh the likely benefits.

Thumb and finger metacarpophalangeal (MP) joint bony avulsion injuries would appear to need surgical stabilization to avoid instability. Many techniques have been described not least to prevent nonunion, but that does not seem to affect the outcome; the results of nonoperative treatment are so good that surgery is rarely required.

The treatment of bone mallet injuries and base of middle phalanx fractures appears not to be primarily to restore bone alignment but to restore gliding as opposed to pivoting. This is established for bony mallet injuries and suggested for proximal interphalangeal (PIP) joint injuries. Extension stress testing of bony mallet injuries and flexing PIP joint injuries appears reliable in predicting the outcome and so the need for surgery. For some of these injuries, there may be a small subset which would benefit from surgery. These need to be identified to improve treatment and to reconcile the strong and conflicting views of surgeons.

Keywords: hand fractures, metacarpal, spiral, transverse, avulsion, mallet, gliding, pivoting

3.1 Introduction

Historically, most hand fractures were treated nonoperatively (conservatively). More recently surgical treatment of many hand fractures has become popular. This involves closed or open reduction and stabilization with Kirschner (K) wires or internal fixation, the latter led by the AO group.1 Although closed K-wiring or open reduction and internal fixation (ORIF) and their many variants can be used to treat almost any hand fracture, it does not mean they should. Clinicians and patients can easily be misled by the purported advantages of surgery. Case series reporting good results may be presented or published and surgeons may be persuaded to follow the new technique despite limited information about its efficacy and particularly the risks. Yet most hand fractures do well with nonoperative treatment.² Even relatively unsuccessful surgical treatment, for example, malunion following K-wiring or ORIF, may give good subjective and objective outcome despite the operation.

The choice to recommend surgery or not will depend on many factors not just objective medical evidence. Training, surgeon and patient beliefs about the efficacy of surgery, remuneration, targets, workload, ignorance, bias related to a recent case, and other factors will influence how clinicians guide patients and how they respond. Ideally, the advice we give should be fully informed and objective, but it is given by humans to humans, so biased by many factors; this is part of the art of medicine. Nonetheless, the advice should be supported by the best available evidence although that is usually suboptimal. Above all, I believe we should know how well a particular injury responds to nonoperative treatment before trying to improve that with surgery because of the increased risks and costs to society and the individual.

The aim of this review is to identify the fractures which can still be best treated nonoperatively. A secondary aim is to try to establish where there may be subgroups of those fractures which might benefit from different treatment such as surgery.

3.2 Methods and Materials

Multiple electronic and subsequent hand searches of published literature were performed in 2015 to identify fractures that do so well with nonoperative treatment that it is currently too difficult for surgery to provide significant further benefit. This review of the literature was updated in 2017. The inclusion criteria were all adult (ε 16 years of age) fractures of the hand, that is, excluding carpal injuries; papers which had a minimum of five cases; and papers with a minimum follow-up of 2 years. This proved impractical as so few had such adequate follow-up and so all papers with a minimum follow-up of 6 weeks were included. Even then reports often had patients younger than 16 years who could not be separated out. The exclusion criteria were fractures where surgery is plainly needed and proven to have been beneficial: many open fractures; replantations; many crush injuries ( Fig. 3.1); displaced intra-articular fractures (excluding bony mallet injuries and ligament avulsion fractures); and pediatric fractures.

3.3 Results

Fingers

3.3.1 Spiral or Long Oblique Metacarpal Fractures

The spiral or long oblique fractures of the metacarpals are very common ( Fig. 3.2). They can be treated surgically; various techniques have been described and good results reported.3–5 Similar good results have been reported with nonoperative treatment.⁶ Khan and Giddins (2015)⁷ showed that all spiral metacarpal fractures, even in the presence of initial malrotation, can be treated nonoperatively with excellent or very good outcomes and minimal morbidity ( Table 3.1). This also applies to double spiral fractures and probably to triple spiral fractures, but they are rare, so the data are not robust. All the patients in their study were treated with early mobilization, without a splint or plaster. They were encouraged to “make a fist” at the first-hand clinic visit in order to correct any malrotation and ensure early mobilization. Twenty-five of 30 patients were reviewed at a minimum of 6 months following their injury. Objectively, they had full, painless movement, and grip strength of at least 90% of the other (uninjured) hand. The only adverse outcomes were minimal malrotation in one patient and mild discomfort in another.

Other authors have highlighted the risk of hand dysfunction caused by shortening of the metacarpals following nonoperative treatment;^8–10 a biomechanical study in 2013 suggested that shortening up to 5 mm is not biomechanically significant.¹¹ This fits with the results of the study of Khan and Giddins. It is confirmed in humans and cadavers that the metacarpals derotate through tightening of the intermetacarpal ligaments. This limits shortening to 2 to 5 mm. Their results further confirm that shortening of 2 to 5 mm is not typically clinically relevant.

Fig. 3.1 Comminuted crush fracture of the thumb.

Malrotation following spiral metacarpal fractures almost always corrects with finger flexion. If it does not, then encouragement or (very rarely) manipulation under local anesthetic would be appropriate to avoid of rotational malunion. Nonoperative treatment gives such good results that recommending surgery seems unjustifiable in almost all patients.

Recently, I have given some patients wrist splints for comfort for the first 1 to 2 weeks from injury. My concern is in restricting full finger flexion, but I suspect that is not a problem and some patients are very sore for the first week or two following these fractures.

3.3.2 Finger Transverse Metacarpal Fractures—Shaft and Neck

In the past, virtually all patients with transverse metacarpal shaft ( Fig. 3.3a, b) and particularly neck (boxer’s) fractures were mobilized freely; the fractures healed with some deformity but almost always with good function. In the United Kingdom, Barton² showed the benefits of a small plaster or splint support to reduce and successfully maintain an improved angulation for transverse metacarpal shaft fractures with very good functional results. More recently the results of surgical treatments have been reported with equally good outcomes.

Fig. 3.2 (a, b) Spiral metacarpal fracture.

Fig. 3.3 (a, b) Transverse metacarpal shaft fracture.

A crucial question is what degree of metacarpal mal-union is acceptable? The answer is not clear. For metacarpal neck (boxer’s) fractures, it has been suggested by various authorsas50to60degrees 12; 30 degrees^13,14; and 20 degrees of flexion^15,16; of note, the amount of acceptable angulation has increased over time. For little finger metacarpal shaft fractures, 30 degrees^12,17 has been suggested as acceptable. These recommendations are only expert opinion.

The outcome of nonoperative treatment has been reported widely and apart from a mild cosmetic abnormality, there is typically an excellent functional outcome (Cochrane review 2005 [updated 2009]).¹⁸ The Cochrane review also noted that there is no good evidence that more marked malunion reduces hand function or gives an unacceptable deformity.¹⁸

Many surgical techniques have been described including intramedullary nailing,¹⁹ intra-medullary screw fixation,^20–22 K-wire (bouquet) fixation,^23–25 intraosseous loop wire fixation,³ and external fixation.²⁶ The results of these techniques are not reliably better than nonoperative treatment and they introduce complications not seen with nonoperative treatment. Nonoperative treatment has complications but apart from malunion these are rare.

Comparative studies, particularly randomized controlled studies (RCTs) provide the best way to assess different treatments. Excellent RCTs in hand surgery are rare, but there are some useful comparative studies. Westbrook et al²⁷ compared the nonsurgical and surgical treatment of metacarpal neck and shaft fractures in a retrospective study. About 105 metacarpal neck fractures were treated nonoperatively, compared with 18 treated operatively (intramedullary K-wiring in 13 and plating in 5 cases); and 113 metacarpal shaft fractures were treated nonoperatively compared with 26 treated operatively (K-wiring in 4 and plating in 22 cases). At a minimum follow-up of 2 years, there were no differences in DASH score, grip strengths, or aesthetics between the nonoperative and operative groups. However, there was a significantly higher complication rate in the patients treated operatively compared to those treated nonoperatively. The follow-up rates were low (17% for nonoperative treatment and 54% for operative treatment); this is typical in these patient groups. A randomized study of metacarpal neck fractures by Strub et al²⁸ has suggested that surgery may give very slightly better outcomes than nonoperative treatment, primarily better cosmesis due to less malunion. They studied two groups each of 20 patients who were pseudo-randomized between nonoperative treatment and intramedullary (bouquet) wiring. The latter required a minimum of two operations each: one to insert and one to remove the wires. The only complications were in the operative group; they also had more dissatisfied as well as very satisfied patients. This study did not measure the inconvenience and costs to the patient or the health care system, so the cost of this possible small benefit is unknown. Of particular note in this and many other studies of surgery for hand fractures, it is not clear how many patients need to be improved from “satisfied” to “very satisfied” to compensate for one “dissatisfied” patient.

Overall, there may be a small cosmetic benefit from surgery for transverse metacarpal shaft and neck fractures of the ring and little fingers, but the costs and the risks are probably not worth the small potential benefit to most patients and in particular most health care systems. The index and middle fingers tolerate transverse metacarpal shaft and neck fracture malunion less well because of the stiffness of the carpometacarpal (CMC) joints, so the indications for surgery are greater. However, only so few studies have been reported that it is not possible to provide a reliable conclusion about their best treatment.

3.3.3 Finger Proximal Phalanx Collateral Ligament Avulsion Fractures

Bekler et al 29 stated that avulsion fractures of the bases of the phalanges of the fingers are challenging to treat ( Fig. 3.4); they also stated that “Avulsion fractures (of the bases of the phalanges) are intra-articular according to their configuration and need anatomical reduction.” There was no evidence to support this opinion. Despite this assertion, finger proximal phalanx avulsion fractures attract a full range of “proven” treatment advice.

Fig. 3.4 Index finger ulnar collateral ligament bony avulsion injury.

Many authors have recommended that all base of finger avulsion fractures should be treated surgically because of the high rate of symptomatic non- or delayed union.30–35 In particular Shewring and Thomas³⁵ reported that of eight of their patients treated nonoperatively all had delayed union; seven were treated with ORIF and bone grafting. They report very good results at discharge from follow-up at 3 months following surgery. In contrast, Sawant et al³⁶ reported that finger proximal phalanx avulsion fractures of up to 25% of the articular surface on the posteroanterior (PA) radiograph achieve very good results with early protected mobilization, that is, with buddy strapping and avoidance of heavy activity for 4 to 6 weeks. Even that level of support may not be necessary or not for so long, but that has not yet been established. Overall, it appears that for avulsion fractures protected mobilization gives results that cannot easily be improved by surgery ( Table 3.2). The data are, however, limited; these are small series or have limited follow-up. The dichotomy with the experience of Shewring and Thomas³⁵ who reported symptomatic delayed union in eight consecutive patients and the excellent results of Sawant et al³⁶ may be because many of these injuries do not unite with bone (as for thumb MP joint ulnar collateral ligament [UCL] avulsions³⁷) but heal with sufficient stability that surgery is not required. This seems to be true for the central as much as for the border digits including the radial collateral ligament (RCL) insertion to the index finger.

As so often in the hand bone union appears much less important than elsewhere in the long bones, but it is such an obvious measure to assess that it can encourage treatment that is not functionally justified.

3.3.4 Bony Mallet Injuries

There are multiple papers reporting on a myriad of techniques for reducing and holding the dorsal avulsion fracture fragment in bony mallet injuries.38–51 Yet, this is an operation with an acknowledged high risk of complications.^30,52,53 Authors of more recent techniques report fewer complications particularly skin breakdown.^38–51 In general, good results are reported for various surgical treatments of bony mallet injuries with a dorsal fracture fragment of one-third or more ( Table 3.3). A number of authors, especially Stark et al,⁵⁴ have recommended surgery to treat fractures of one-third or more. These authors have been concerned about achieving “anatomical reduction” of the fracture and prevention of subluxation of the main distal phalanx fracture fragment.^39–51 How much subluxation needs to be treated is unproven, but it is known that some cases do progress to symptomatic dislocation. In one of the most widely cited papers, Wehbe and Schneider⁵³ reported that among patients with dorsal fracture fragments of over one-third followed up for a mean of 3.2 years, the 15 patients treated nonoperatively did as well as the six treated operatively. They concluded that operative treatment gave no better results than nonoperative treatment. Similarly, another series of cases with a 5-year follow-up has shown that both tendinous and bony mallet injuries treated nonoperatively achieve good objective and very good subjective outcomes.55 Among the bony mallet injuries were some with fracture fragments greater than one-third, but they are not reported separately. They also reported some evidence of degeneration at the distal interphalangeal (DIP) joint in 10 of 11 patients with large fracture fragments within 5 years. Wehbe and Schneider⁵³ noted no difference in the rates of radiographic DIP joint osteoarthritis (OA) between operative and nonoperative treatment. Other authors have reported rates up to 50%, yet some reported 0% ( Table 3.3). Almost certainly their criteria (which are reported rarely) differ, making comparison difficult. The risk of radiographic OA would be a potential concern except that long-term symptomatic degenerative arthritis in the DIP joints is not widely reported in patients who have had bony mallet injuries, that is, as hand surgeons we rarely see patients requiring treatment for symptomatic DIP joint arthritis who had bony mallet injuries decades earlier. It would appear that the majority of patients with bony mallet injuries do not require surgical treatment.

A Cochrane review of bony mallet injuries reported that there was a paucity of good studies and there was no evidence that surgery was better than nonoperative (typically splint) treatment for all types of mallet injury.56 They did, however, acknowledge that there may be a subgroup of these injuries that would benefit from surgery.

The key concern is volar subluxation of the bulk of the distal phalanx ( Fig. 3.5). Several recent papers have clarified further the risk of subluxation: Kim and Kim⁵⁷ showed that over 50% of patients with a fracture fragment involving greater than one-third of the base on the lateral radiograph did not progress to subluxation. They noted a cutoff of at a fracture fragment size of 48% of the base, above which subluxation was more likely. This is also similar to the findings in the study of Giddins⁵⁸ who noted that stable injuries had a mean fracture fragment size of 49%. Kim and Kim⁵⁷ also related subluxation to delay in application of a splint, particularly to a delay of longer than 12 days. Moradi et al⁵⁹ reviewed 392 bony mallet injuries. They showed that subluxation did not occur with a fracture fragment smaller than 39% on the lateral radiograph. The risk of subluxation increased with increasing size of the fragment, increased displacement of the fragment, and time from injury to treatment. Nonetheless, most (68%) did not sublux even with a fracture fragment greater than 39%. Giddins 58 has shown that the risk of subluxation can be assessed by a lateral hyperextension radiograph performed within 1 to 2 weeks of injury. If there is gliding of the joint ( Fig. 3.6), that is, it remains congruent as it rotates into extension, then this is a stable joint. The presumption is that there has not been so much volar plate and collateral ligament injury that subluxation will occur. If there is pivoting ( Fig. 3.7), then subluxation will usually occur, although this may only be mild. A third subgroup occurs described as tilting ( Fig. 3.8). This appears to be a variant on gliding and is not typically associated with subluxation. Exactly what level of subluxation requires treatment is unclear.