Predictors of unstable distal radius fractures: a systematic review and meta-analysis

Abstract

The aim of this study was to perform a systematic review in order to identify predictors of secondary displacement in distal radius fractures. We performed a systematic review and identified all studies that reported secondary displacement following distal radius fractures. Where possible, we pooled the odds ratios of predictors. The initial search yielded 3178 studies of which 27 were included. Multiple studies found that age, shortening, volar comminution, loss of radial inclination, the presence of a volar hook, AO type 3 fractures (A3, B3, C3) and the Older classification were significant predictors of secondary displacement. Pooling revealed a significantly increased risk of secondary displacement in fractures with dorsal comminution, in women and in patients aged >60 years. An associated ulna fracture or intra-articular involvement does not result in an increased risk of secondary displacement. The overview provided in this study can help surgeons to inform patients of the chances of success of closed treatment regarding the radiological outcome and facilitate shared decision making.

Level of evidence: II

Keywords

Distal radius fracture instability displacement predictors loss or reduction loss of position unstable

Introduction

Initial treatment of patients with distal radius fractures generally consists of closed reduction and plaster immobilization (Arora et al., 2012; Earnshaw et al., 2002; Kumar et al., 2008; Mackenney et al., 2006). However, fracture re-displacement following closed reduction occurs in up to 64% of the patients (in part depending upon the definition) (Jenkins, 1989; Mackenney et al., 2006; Makhni et al., 2008). Most surgeons would agree that distal radius fractures with re-displacement outside acceptable parameters (>10° dorsal angulation, radial shortening >3 mm or intra-articular step-off) (American Academy of Orthopaedic Surgeons Board of Directors, 5 December 2009) benefit from surgical fixation, taking into account patient-related factors.

Patients with a distal radius fracture with a perceived high risk of re-displacement may be recommended primary surgical treatment. Unfortunately, patients with these potentially more unstable distal radius fractures are difficult to identify.

In 1989, Lafontaine et al. identified five factors predictive of fracture instability: dorsal angulation exceeding 20° at presentation; dorsal comminution; extension of the fracture into the radiocarpal joint; an associated ulnar fracture; and age over 60 years (Lafontaine et al., 1989). According to Lafontaine et al. (1989), a fracture can be considered potentially unstable if three or more factors are present.

Since then, several studies have confirmed and refuted the importance of these five risk factors, and new clinical and radiological predictors have been identified (Abbaszadegan et al., 1989; Nesbitt et al., 2004; Poigenfürst, 1980; Tahririan et al., 2013). Other authors have quantified predictors and presented scoring systems to predict the risk of secondary displacement based on clinical and radiological variables (Adolphson et al., 1993; Mackenney et al., 2006).

Although several risk factors of secondary displacement are commonly accepted, the evidence for some predictors is limited. An overview of all known predictors of fracture instability in literature, and their relative weight (pooled odds ratios), could assist physicians in decision making regarding the optimal method of treatment for patients with distal radius fractures.

The aim of this study was to perform a systematic review in order to identify predictors of secondary displacement in distal radius fractures.

Methods

This review was conducted and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al., 2010). A review protocol was drafted and registered on PROSPERO with number CRD42014010828. All of the following steps were performed by two independent reviewers (SA and MMJW). Disagreements between the two reviewers were discussed until a consensus was reached.

Eligibility criteria systematic review

Any study that could potentially provide information on predictors of secondary displacement was eligible for inclusion. Therefore, studies describing non-operative treatment in patients with distal radius fractures were also included. We anticipated that these studies would report secondary displacement during follow-up.

Inclusion criteria were studies that addressed adult (18 years and older) patients with distal radius fractures initially treated with plaster immobilization, with or without closed reduction of the fracture. Reviews, animal studies, cadaver studies, case reports, surveys, current (management) concepts, editorials, commentaries, conference abstracts and letters were excluded. Since there are numerous definitions of displacement and instability, we did not predetermine a definition that should have been used for a study to be included. For the same reason we did not define what constitutes a displaced, undisplaced or minimally displaced fracture.

Eligibility criteria meta-analysis

For the meta-analysis, an additional inclusion criterion was studies that reported odds ratios of predictors of secondary displacement, or provided sufficient information to calculate odds ratios. We required studies to have had a minimal radiological follow-up of 4 weeks for the assessment of secondary displacement.

Literature search and study selection

We conducted a systematic search of the MEDLINE and EMBASE databases on 28 August 2013, and updated the search on 19 March 2015 (Appendix 1, available online), to identify all studies on patients with distal radius fractures and reporting displacement or fracture instability. To ensure proper interpretation of the results by our team, publication language was restricted to English, German, Spanish, French, Turkish or Dutch. The resulting titles and abstracts were reviewed. If the eligibility criteria were met, full manuscripts were procured and reviewed. The bibliographies of all included studies were manually screened for additional articles of interest.

Data extraction

Data were extracted using a standardized data collection form that was developed according to the Cochrane guidelines (Higgins and Green, 2011). Items collected included publication details, study type, numbers of patients, fracture types, the definition of displacement, the type of statistical analysis and details regarding treatment. For each study we determined which predictors of secondary displacement were found statistically significant (p < 0.05). Additionally, we assessed if the predictor was tested in a univariate or a multivariable analysis. For the meta-analysis, predictors and odds ratios or coefficients of predictors of secondary displacement were obtained and collected in a database.

Quality assessment

Quality and heterogeneity across studies was assessed using the Quality in Prognostic Studies (QUIPS) tool (Hayden et al., 2013; Huguet et al., 2013) (Table 1 and Appendix 2, available online). The QUIPS tool rates individual studies according to the potential risk of bias associated with six domains: (1) study participation; (2) study attrition; (3) prognostic factor measurement; (4) outcome measurement; (5) confounding measurement and account; and (6) analysis. Two reviewers (SA and MMJW) independently assessed the potential risk of basis of each study based on information available in the manuscripts. Disagreements were discussed by the reviewers until a consensus was reached.

Table 1.

Quality assessment according to QUIPS tool.

First author	Bias due to study participation	Bias due to study attrition	Bias due to prognostic factor measurement	Bias due to confounding	Bias due to analysis	Bias due to outcome measurement
Abbaszadegan	Moderate	Moderate	Low	Low	Moderate	Moderate
Adolphson	Moderate	Moderate	Low	Low	Moderate	Moderate
Alemdaroglu	Moderate	Low	Low	Low	Moderate	Low
Altissimi	Moderate	Moderate	Moderate	Moderate	Moderate	High
Bartl	Low	Low	Moderate	Moderate	Moderate	High
Benoist	NA	NA	NA	High	NA	NA
Beumer	Moderate	Low	Moderate	Moderate	Moderate	Moderate
Blakeney	Low	Low	Low	Moderate	Moderate	Moderate
Camelot	Low	Low	Low	Low	Moderate	Low
Clayton	Moderate	Low	Low	Low	Low	High
Einsiedel	High	Not reported	Low	Low	Moderate	Low
Fenyo	Moderate	Not reported	High	High	High	High
Hove	High	Not reported	Low	Low	Low	Low
Kulej	Moderate	Not reported	Moderate	Moderate	Moderate	High
Lafontaine	Moderate	Low	Low	Low	Moderate	High
LaMartina	Low	Low	Low	Low	Moderate	Low
Leone	Low	Not reported	Moderate	Low	Moderate	Low
Mackenney	Low	Low	Low	Low	Low	Low
Makhni	Moderate	Not reported	Low	Low	Moderate	Low
Myderrizi	Moderate	Not reported	Moderate	Low	Moderate	High
Nesbitt	Moderate	Low	Low	Low	Low	Low
Oskarsson	Low	Low	Moderate	High	High	High
Phillips	Moderate	Not reported	Moderate	Low	Moderate	High
Porter	Low	Low	Low	Moderate	High	Moderate
Robin	Moderate	Low	Moderate	Moderate	Moderate	Moderate
Tahririan	Moderate	Low	Moderate	Low	Low	Low
Wadsten	High	Low	Moderate	Moderate	Low	Moderate

Synthesis of results of meta-analysis

For the meta-analysis, univariate odds ratios were pooled using the random effects model. The random effects model involves an assumption that the effects being estimated in the different studies are not identical, but follow some distribution (Higgins and Green, 2011). If coefficients from a logistic regression analysis were reported, odds ratios were calculated by taking the exponential. Odds ratios for instability or displacement at any point during follow-up (early or late) were added as separate odds ratios to the analysis. Data were pooled with the log odds ratio and the standard error with the generic inverse variance.

To test the robustness of the results, sensitivity analyses were performed by comparing the results obtained with the random effects model with the fixed effects models (Higgins and Green, 2011). Similar results imply that the results are robust.

Statistical heterogeneity for each analysis was tested with I². The I² is a statistic that indicates the percentage of variance and is qualified as follows: 0% to 40% (might not be important); 30% to 60% (may represent moderate heterogeneity); 50% to 90% (may represent substantial heterogeneity); 75% to 100% (considerable heterogeneity) (Higgins and Green, 2011).

Results

Search results

The initial search yielded a total of 3178 studies, of which 2249 remained after excluding the duplicates. The full text manuscripts of 133 studies were evaluated and 27 studies were included in the systematic review (Figure 1) (Abbaszadegan et al., 1989; Adolphson et al., 1993; Alemdaroglu et al., 2010; Altissimi et al., 1994; Bartl et al., 2014; Benoist and Freeland, 1995; Beumer and McQueen, 2003; Blakeney and Webber, 2009; Camelot et al., 1998; Clayton et al., 2009; Einsiedel et al., 2009; Fenyo and Johansson, 1974; Hove et al., 1994; Kulej et al., 2007; Lafontaine et al., 1989; LaMartina et al., 2015; Leone et al., 2004; Mackenney et al., 2006; Makhni et al., 2008; Myderrizi, 2011; Nesbitt et al., 2004; Oskarsson et al., 1997; Phillips and Al-Shawi, 2014; Porter and Stockley, 1987; Robin et al., 2014; Tahririan et al., 2013; Wadsten et al., 2014). The full text manuscript of one study could not be procured, even after contacting the authors (Plaweski S et al., 1997).

Figure 1.

Flowchart.

Systematic review

The characteristics of the studies included are outlined in Table 2. All but three studies focused primarily on secondary displacement or radiographic outcome (Bartl et al., 2014; Oskarsson et al., 1997; Porter and Stockley, 1987). One study excluded fractures that lost reduction within 1 week and assessed late secondary displacement at 2 months (Myderrizi, 2011). Quality and risk of bias were variable (Table 1 and Appendix 2, available online).

Table 2.

Characteristics of studies included (N = 27).

First author	N	Study population	Age (years)	Primary outcome	Treatment	Duration of follow-up	Patients with instability (%)	Predictors tested	Predictors found significant (p < 0.05)
Abbaszadegan	263	Colles fractures	Mean: 62 Range: 18–97	Anatomic failure at 8 weeks (RS <4 mm, DA <30°) or DRFs that were re-reduced within first 2 weeks	Closed reduction and dorsal plaster 4 weeks	8 weeks	NR	Age, initial RS, Lidström, Frykman, dorsal compression, radial angle, comminution	Age, initial RS Lidström, dorsal compression, comminution
Adolphson	107	Displaced Colles fractures	Mean: 67 Range: 51–86	Probability of good anatomical end result at 8 weeks (RS <4 mm, DA <30°)	Closed reduction and dorsal plaster splint 4 weeks	8 weeks	61 (57)	Age, initial RS, Lidström classification combined as a prediction model	Age, initial RS, Lidström classification combined as a prediction model
Alemdaroglu	75	Displaced extra-articular DRFs	Mean: 53 Range: 24–83	Redisplacement (increase of 10° of DA or VT, or loss of SD 5° RD or >2 mm increase in UV or a combination of 5° degrees of DA and 3° of RA and 1 mm of RS) any time during follow-up	Closed reduction and short arm plaster with wrist in 10°–15° of flexion; the degree of UD was at surgeon’s preference but limited to 0°–30°	[not specified] Mean: 9 months	24 (32)	Age, sex, initial RI, initial VT, initial UV, DC, alignment-related indices, cast-related indices, extent of UD of cast, non-anatomical reduction, co-existing ulnar fracture, obliquity of fracture line	Casting technique, degree of the UD of the cast, obliquity of the fracture line, inadequate reduction
Altissimi	97	Dorsally displaced DRFs	Mean: 55	Secondary displacement [undefined]	Closed reduction and above elbow cast 5 weeks with wrist in moderate palmar flexion and UD	5 weeks	71 (73)	Initial RS, DA, RA, Older’s type	Initial shortening, Older’s type 3 and 4
Bartl	88	Patients >65 with unstable intra-articular DRF (AO types C1, C2, C3)	Mean: 74.4 SD: 7.1	Significant loss of reduction or pronounced joint incongruence. [undefined]	Closed reduction and closed forearm cast for 6 weeks	1 year	37 (42)	AO types C1, C2 and C3	AO type C3
Benoist	0	NA	NA	Unstable DRFs	NA	NA	NA	Shelf sign	NA
Beumer	59	Patients >60, low-demand or demented, with DRF	Mean: 82 Range: 65–93	Maintenance of primary reduction (VT or DA <10°, RS <3 mm, and carpal alignment	Closed reduction	NR	50 (85)	AO fracture classification, initial displacement	-
Blakeney	162	Minimally displaced (48) and displaced (114) extra-articular Colles fractures	Mean: 58 SD: 19.6	Poor radiological results (DA ⩾10° or VT ⩾20°) and/or progression to surgery	Closed reduction and plaster-of-Paris back-slab	6 weeks	77 (48)	DA >15°, inadequate reduction	DA >15°, inadequate reduction
Camelot	189	DRFs	Mean: 53.6 Range: 15–94	Secondary frontal displacement (>20° RI), sagittal displacement (DA > 0°) or axial displacement (RS >5 mm)	Closed reduction according to Judet and circular plaster with wrist in flexion and 45° UD	42 days	169 (89)	Age, sex, ulnar styloid fracture, RS, DC, RI, volar hook, UD of plaster	Secondary frontal displacement: age >55, DA, RS >4 mm, UD plaster <30°. Secondary axial displacement: age >55 and DC Secondary sagittal displacement: RS >4 mm and volar hook
Clayton	137	Patients aged >55 with low energy displaced and non-displaced DRFs	Mean: 71 Range: 45–87	Early (<2 weeks) instability. Calculated with previously published algorithm	Closed reduction for displaced fractures and plaster-of-Paris back-slab	6 weeks	Early: 45 (33) Late: 37 (27)	Osteoporosis (T-score)	Osteoporosis (T-score)
Einsiedel	311	Conservatively treated DRFs	Median: 68 Range: 15–100	DA at healing	NR	NR	NA	DA, RA, RS	DA, RA, RS
Fenyo	100	DRFs where primary reduction was indicated and completely controlled	NR	Secondary displacement (DA or VT >10°)	Closed reduction and dorsal plaster splint. In 2/3 of the cases completed by a circular plaster after 7–14 days	40 days	20 (20)	Result of primary reduction, immobilization technique (dorsal plaster splint or circular)	-
Hove	645	Conservatively treated displaced (446) and non-displaced (199) Colles fractures	Mean: 61 SD: 16	Anatomical end result (radiographic DA and radial length)	Closed reduction (for displaced fractures) and below elbow dorsal cast with 3-point pressure and wrist in neutral extension for 4 weeks	Until union	NA	Age, sex, energy, energy of trauma (high or low impact), initial DA, initial radial length, Older’s classification	DA at union: initial DA, age, Older type Radial length at union: age, initial radial length, initial DA
Kulej	395	Conservatively treated DRFs	Mean: 59 Range: 19-84	Maintenance of surgical outcome following closed reduction [undefined]	Closed reduction and plaster immobilization (mean 40 days)	624 days (mean)	NR	AO types	AO types A2, C2, C3
Lafontaine	112	DRFs	Mean: 54 Range: 13-93	Poor radiologic outcome based on Stewart Score	Closed reduction and plaster-of-Paris cast	6 weeks	42 (38)	Age >60, DA >20°, DC, intra-articular radiocarpal involvement, associated ulnar fracture	Age >60, DA >20°, DC, intra-articular radiocarpal involvement, associated ulnar fracture
LaMartina	168	Complete metaphyseal (AO type A2, A3, C1, C2) DRFs with ⩾10° initial DA		Change in DA, RH, UV and RI and DA, RH, UV and RI at union	Closed reduction and 3-point short-arm circumferential plaster cast with the wrist in slight flexion and UD	6 weeks or until healing	NA	Age, DC, volarhook, and intra-articular fracture	Change in DA: volar hook, DCChange in radial height: ageChange in UV: DCChange in RI: ageDA at union: volar hook, DCUV at union: ageRH at union: age, intra-articular fractureRI at union: age
Leone	71	Dorsally displaced (N = 50) and non-displaced (N = 21) extra-articular DRFs	Mean: 65	Early (<1 week) and late (<6 weeks) instability, defined as DA>15°, VT >20°, UV >4 mm, RI <10°	Closed reduction (for displaced fractures) and plaster cast with wrist in neutral rotation, UD and slight flexion for 6 weeks	6 weeks	Early: 14 (20)Late: 24 (34)	Age, sex, RS, DA, DC, RI, VT, ulnar styloid fracture	Early: RS and (negative) VTLate: age, RI, RS and VT
Mackenney	3081	Dorsally displaced (1595) and minimally displaced (N = 1486) extra-articular DRFs	Mean: 64Range: 14–100	Early (2 weeks), and late (6 weeks) instability (DA >10° and/or UV > 3 mm)	Closed reduction (for displaced fractures) and dorsal plaster-of-Paris slab	6 weeks	Early in minimally displaced: 149 (10)Late in minimally displaced: 244 (22)Early in displaced: 682 (43)Late in displaced: 391 (47)	Age, sex, mechanism of injury, independence, type of comminution, AO/OTA type, AO/OTA group, AO/OTA subgroup, Frykman classification, initial DA, initial radial shift, initial UV, DA at 1 week, radial shift at one week, UV at 1 week	Early in minimally displaced: age, any form of comminution, AO/OTA type 3, initial DA, UVLate in minimally displaced: age, comminution, DA at 1 week, UV at 1 weekEarly in displaced: age, initial UV, any type of comminutionLate in displaced: Age, DA at 1 week, UV at 1 week
Makhni	124	DRFs treated with cast (N = 50) and treated with closed reduction and cast (N = 74)	Mean: 56	Secondary displacement that necessitated operative fixation or loss of reduction at union (DA <10°, VT<25°, RS <5 mm, articular gap/step-off <2 mm)	Closed reduction (for displaced fractures) and cast immobilization in sugar-tong splint	8 weeks	79 (64)	Age, sex, comminution	Age
Myderrizi	329	Conservatively treated DRFs with anatomical reduction and without loss of reduction within 1 week	Mean: 54.3Range: 18–81	Late collapse at 2 months [undefined]	Manipulation and initial immobilization with a cast splint and later a below elbow full arm cast	2 months	171 (52)	Age, sex, fracture type (AO), DRUJ involvement, ulnar styloid fracture, initial displacement (VT, radial tilt, UV)	Age >70, AO classification, DRUJ instability in combination with ulnar styloid fracture, initial displacement
Nesbitt	50	Displaced DRFs with acceptable reduction and ⩾3 Lafontaine’s instability factors	Mean: 60Range: 16–87	Acceptable radiographic result at 4 weeks (RS ⩾2 mm, change in RI ⩾5°, step-off ⩾2 mm, loss of VT ⩾10°)	Closed reduction with finger trap traction and sugar-tong splint	4 weeks	27 (54)	Age, age >60, DA, DA >20°, comminution, intra-articular involvement, associated ulnar fracture, RS	Age
Oskarsson	158	Colles fracture	Range: 25–75	Poor functional outcome	Closed reduction when needed and cast for 4 to 6 weeks	8 weeks	NR	Immobilization period, Frykman type, associated ulnar styloid fracture	Frykman type 4 with ulnar styloid fracture (p-values not reported)
Phillips	87	Consecutive manipulated DRFs and with <2 mm intra-articular displacement	Mean: 62.8SD: 13.8	Early instability (RS >4 mm, DA >10° or intra-articular displacement or step-off >2 mm)	Closed reduction and immobilization in a plaster-of-Paris castback, completed with Scotchcast	6 weeks	NR	Age, volar cortex restored, initial DA, fracture ulnar styloid or ulnar head, intra-articular fracture, DC	Age, volar cortex restored and ulnar styloid or ulnar head fracture
Porter	115	DRFS with acceptable position (50) and displaced DRFs (65)	Average: 69	Functional outcome Redisplacement defined as: change in RH >3 mm, and/or change in DA 10–20°	Closed reduction and short arm back slab with wrist in moderate flexion and UD, average 30 days	2 years	67 (58)	Poor reduction, comminution, intra-articular involvement	-
Robin	78	Patients >65 with displaced DRF and a DEXA scan performed within 1 year	Mean: 78Range: 65–94	Loss of reduction expressed as percentage of retained reduction in VT, RH, RI, UV*	Closed reduction with 3-point moulding and after 3–4 weeks left in sugar tong splint or converted to short-arm cast	11 weeks	Loss in VT: 7 (9)Loss in RH: 41 (53)Loss in RI: 34 (44)Loss in UV: 42 (54)	Osteoporosis (T-score of lumbar spine and femoral neck)	-
Tahririan	157	Displaced DRFs with acceptable reduction and ⩾3 Lafontaine’s instability factors	Mean: 51Range: 20–87	Failure to maintain acceptable reduction (RS <2 mm, change in RI <5°, step-off <2 mm, loss of VT <10°)	Closed reduction and immobilization by a forearm plaster-of-Paris cast with moderate volar tilt and UD	6 weeks	92 (59)	Age, sex, RH, loss of RH, DC, ulna fracture, RI, loss of RI, DA, intra-articular involvement	Loss of RH, loss of RI, age
Wadsten	387	DRFs with minimal displacement (VT <20°, DA <10°, RT >10°, UV <2 mm and intra-articularstep off <2 mm) and displaced DRFs (beyond boundaries of minimally displaced)	Mean: 56Range: 15–74	Redisplacement [undefined but seems to be the same definition as for displaced fractures: VT ⩾20°, DA ⩾10°, RT ⩽10°, UV ⩾2 mm and intra-articular step off ⩾2 mm]	Minimally displaced fractures: dorso-radial plaster-of-Paris forearm cast with good 3-point mouldingDisplaced fractures: closed reductionand plaster-of-Paris slab	Until occurrence of displacement or union at 3 months	Overall: 256 (64) Minimally displaced: 53 (32) Displaced: 203 (87)	DC, volar comminution, initial presentation [undefined] and intra-articular involvement	Volar comminution, DC, initial presentation

DA: dorsal angulation; DC: dorsal comminution; DRF: distal radius fracture; DRUJ: distal radioulnar joint; NA: not applicable; NR: not reported; RD: radial deviation; RA: radial angulation; RH: radial height; RI: radial inclination; RS: radial shortening; RT: radial tilt; SD: standard deviation; UD: ulnar deviation; UV: ulnar variance; VT: volar tilt; AO: arbeitsgemeinschaft für osteosy; OTA: Orthopaedic Trauma Association.

This ratio was calculated as the difference between the follow-up radiographic measurement (Fu) and the radiographic measurement of the initial fracture radiograph (fx) divided by the difference between the reduction radiographic measurement (Re) and the radiographic measurement of the initial fx ([Fu e fx]/[Re e fx]). This ratio was calculated for volar tilt, radial height, radial inclination and ulnar variance.

Numbers (precentages).

A total of 7574 patients were included. Secondary displacement occurred in 10% to 89% of the fractures. A total of 44 different predictors of secondary displacement were reported. An overview of the most common predictors and in which study they were assessed is outlined in Table 3.

Table 3.

Overview results of each study for the most common predictors.

Predictor	Significant^a		Not significant^a
	In univariate analysis	In multivariate analysis	In univariate analysis	In multivariable analysis
Age (continuous)	Abbaszadegan	Hove^c	Alemdaroglu
	Adolphson	Mackenney^f
	Lafontaine	Nesbitt^b
	Leone	Tahririan^e
	Makhni	LaMartina^h
	Phillips
Age >55–65	Camelot	Mackenney^f	Makhni
		Nesbitt^b
Sex (female)	Myderrizi^g	Hove^c	Alemdaroglu	Mackenney^f
			Leone	Tahririan^e
			Makhni
			Camelot
Independence				Mackenney^f
Osteoporosis		Clayton^d	Robin
Mechanism of injury				Mackenney^f
Energy of trauma (low/high)				Hove
Shortening	Abbaszadegan	Hove^c	Nesbitt
	Adolphson	Tahririan^e
	Altisimmi
	Einsiedel
	Leone
	Camelot
Initial ulnar variance		Mackenney^f	Alemdaroglu
Initial dorsal angulation (continuous)	Einsiedel	Hove^c	Altissimi	Nesbitt^b
	Leone	Mackenney^f	Phillips	Tahririan^e
	Camelot
Initial dorsal angulation >20°	Lafontaine			Nesbitt^b
(Dorsal) comminution	Abbaszadegan	Mackenney^f	Alemdaroglu	Nesbitt^b
	Lafontaine	LaMartina^h	Leone	Tahririan^e
	Makhni	Wadstenⁱ	Porter
	Camelot		Phillips
Volar comminution		Wadstenⁱ
Initial radial inclination	Leone		Alemdaroglu
			Camelot
Loss of radial inclination		Tahririan^e
Initial radial shift				Mackenney^f
Intra-articular involvement	Lafontaine	LaMartina^h	Porter	Nesbitt^b
			Phillips	Tahririan^e
				Wadstenⁱ
Associated ulnar styloid fracture	Lafontaine		Alemdaroglu	Nesbitt^b
	Oskarsson		Leone	Tahririan^e
	Myderrizi^g		Camelot
	Phillips
Volar hook/volar cortex restored	Camelot	LaMartina^h
	Phillips
AO classification	Myderrizi^g	Mackenney^f (group 3)	Beumer
	Bartl	Mackenney^f (group 3)
Lidström classification	Abbaszadegan
	Adolphson
Frykman classification	Oskarsson (type 4)		Abbaszadegan	Mackenney^f
Older classification	Altisimmi (type 3 and 4)	Hove^c
Non-anatomic reduction	Alemdaroglu		Porter
	Blakeney
	Fenyo
Immobilization technique	Alemdaroglu (ulnar deviation of cast and three-point index)Camelot (ulnar deviation plaster)		Fenyo (dorsal plaster splint or circular)
Immobilization period			Oskarsson

AO: arbeitsgemeinschaft für osteosynthesefragen.

Significant association with secondary displacement at p-value < 0.05.

Adjusted for age, dorsal angulation, dorsal comminution, intra-articular involvement and ulnar styloid fracture.

Adjusted for initial dorsal angulation, initial radial length, age, older types.

Adjusted for age.

Adjusted for age, radial height, dorsal comminution, ulnar styloid fracture, radial inclination, dorsal angulation, intra-articular involvement and sex.

Adjusted for age, sex, mechanism of injury, independence, type of comminution, AO/OTA type, AO/OTA group, AO/OTA subgroup, Frykman classification, initial dorsal angle, initial radial shift, initial ulnar variance, dorsal angle at 1 week, radial shift at 1 week and ulnar variance at 1 week.

Only late displacement, excluded fractures that displaced at 1 week.

Unclear for which variables was adjusted.

Adjusted for age and sex.

Only a few predictors were also tested in a multivariable analysis, adjusting for other covariates. Of these, age, shortening, volar comminution, loss of radial inclination, the presence of a volar hook, AO type 3 fractures (A3, B3, C3) and the Older classification were found to be significant predictors in one or more studies.

The mechanism of the injury, the forces involved, i.e. the estimated energy of the injury, the independence of a patient (defined as being able to go shopping), initial dorsal angulation >20° from neutral, the initial radial shift, the presence of an associated ulnar styloid fracture and the Frykman classification were revealed not to be significant predictors of secondary displacement in one or more studies that adjusted for other covariates. For gender, initial ulnar variance, comminution and intra-articular involvement, there were studies that found a significantly increased risk of secondary displacement and others that did not.

Meta-analysis

We were able to extract or calculate odds ratios from 11 articles (Alemdaroglu et al., 2010; Bartl et al., 2014; Blakeney and Webber, 2009; Hove et al., 1994; Lafontaine et al., 1989; Leone et al., 2004; Mackenney et al., 2006; Makhni et al., 2008; Myderrizi, 2011; Nesbitt et al., 2004; Tahririan et al., 2013). We contacted the corresponding authors of four additional articles that reported p-values without odds ratios or only adjusted odds ratios to request additional information (Camelot et al., 1998; LaMartina et al., 2015; Phillips and Al-Shawi, 2014; Wadsten et al., 2014). The authors of two studies provided further data (Phillips and Al-Shawi, 2014; Wadsten et al., 2014) so we could extract or calculate odds ratios from 13 studies. Ten of these studies used a comparable definition for secondary displacement and were therefore eligible for pooling (Alemdaroglu et al., 2010; Blakeney and Webber, 2009; Leone et al., 2004; Mackenney et al., 2006; Makhni et al., 2008; Myderrizi, 2011; Nesbitt et al., 2004; Phillips and Al-Shawi., 2014; Tahririan et al., 2013; Wadsten et al., 2014). Myderrizi (2011) excluded patients with loss of reduction at 1 week and this study was therefore not included in the analysis. Wadsten et al. (2014) also analysed the 92 patients who went on to surgery because their post reduction films revealed an unacceptable position, and therefore we did not include this study. From the remaining eight articles, we were able to pool the odds ratios of seven predictors. A total of 3807 patients were analysed.

Female gender, age >60–65 years and dorsal comminution were significant predictors of secondary displacement (Figures 2 –4). An associated ulnar styloid fracture, intra-articular involvement, dorsal angulation >15° from neutral and dorsal angulation >20° from neutral were not significantly associated with secondary displacement (Figure 5 –8).

Figure 2.

Forest plot of comparison female versus male. The four odds ratios from Mackenney et al. (2006) represent early and late instability in minimally displaced and displaced fractures. The odds ratios from Leone et al. (2004) represent early and late instability in all fractures.

Figure 3.

Forest plot of comparison age >60 years versus <60–65 years. Mackenney et al. (2006) represent early and late instability in minimally displaced and displaced fractures.

Figure 4.

Forest plot of comparison dorsal comminution versus no comminution. The four odds ratios from Mackenney et al. (2006) represent early and late instability in minimally displaced and displaced fractures. The odds ratios from Leone et al. (2004) represent early and late instability in all fractures.

Figure 5.

Forest plot of comparison presence of associated ulnar styloid fracture versus intact ulnar styloid. The odds ratios from Leone et al. (2004) represent early and late instability in all fractures.

Figure 6.

Forest plot of comparison intra-articular fracture involvement versus no involvement.

Figure 7.

Forest plot of comparison dorsal angulation >15° from neutral versus dorsal angulation ⩽15° from neutral.

Figure 8.

Forest plot of comparison dorsal angulation >20° from neutral versus dorsal angulation ⩽20° from neutral.

A sensitivity analysis showed similar estimates in the random and fixed effects models (Table 4). The odds ratio of intra-articular involvement and dorsal angulation >15° from neutral were both significant only in the fixed effects model.

Table 4.

Sensitivity analysis.

	Random effects model		Fixed effects model
	OR	p-value	OR	p-value
Female vs. male	1.82	<0.00001	1.78	<0.00001
Age >60 years vs. <60–65 years	2.67	<0.00001	2.66	<0.00001
Dorsal comminution	2.29	<0.00001	2.99	<0.00001
Ulnar styloid fracture	1.00	1.00	0.98	0.93
Intra-articular involvement	0.76	0.53	0.62	0.09
Dorsal angulation >15°	4.07	0.15	5.73	0.001
Dorsal angulation >20°	0.92	0.87	0.92	0.87

OR: odds ratio.

Discussion

This systematic review provides an overview of all predictors of secondary displacement of distal radius fractures in literature. We have demonstrated that not all popular predictors of instability that are persistently used in the literature have indeed been identified as significantly associated with secondary displacement. For instance, several studies that also adjusted for other covariates in their analyses did not find an increased risk of secondary displacement in fractures with an initial dorsal angulation exceeding 20° from neutral and neither in fractures with an associated ulnar styloid fracture.

For other popular risk factors, such as the female gender, dorsal comminution and intra-articular involvement, the results were inconclusive. Some studies with adjusted analyses did find a significant influence on secondary displacement and others did not.

In an attempt to provide a definite answer, we pooled the odds ratios of seven predictors, including gender, age >60–65, dorsal comminution, associated ulnar styloid fracture, intra-articular involvement and dorsal angulation exceeding 15° and 20° from neutral. Our results show a significantly increased risk of secondary displacement in fractures with dorsal comminution and in women. Additionally, the pooled results confirm the importance of age, demonstrating a significantly increased risk of secondary displacement of distal radius fractures in patients older than 60–65 years.

Conversely, our analysis reveals no significantly increased risk of secondary displacement in fractures with a dorsal angulation exceeding 15° or 20° from neutral, an associated ulna fracture or intra-articular involvement.

Intra-articular involvement is often mentioned as an indication for surgery and is one of Lafontaine’s often cited risk factors (American Academy of Orthopaedic Surgeons Board of Directors, 5 December 2009; Mackenney et al., 2006; Makhni et al., 2008). Nevertheless, we found no significant effect on secondary displacement of intra-articular fractures (0.52, p = 0.07). When the fixed effects model was used, risk of secondary displacement in intra-articular fractures was significantly lower (OR 0.5, p = 0.03). An explanation for this is that in the fixed-effects analyses, studies are weighted less equally than in the random effects. Therefore, the larger study of the two (that found a significant OR) has greater weight (Higgins and Green, 2011). Another possible reason could be that patients with more severe intra-articular fractures received primary operative treatment, resulting in less severe intra-articular fractures in the study population. The same reason may explain why dorsal angulation >15° from neutral was significant in the fixed-effects model and not in the random-effects model.

Another of Lafontaine’s risk factors is the presence of an associated ulnar styloid fracture, which is believed to result in injuries of the ulnocarpal ligaments and therefore constitute fracture instability (Poigenfuerst and Tuchmann, 1978; Poigenfürst, 1980). However, despite its popularity, this predictor was not identified as significantly associated with secondary fracture displacement in multiple studies. This is confirmed by the results of our meta-analysis, which do not show an increased risk of secondary displacement for an associated ulnar styloid fracture.

The strength of this study is that it provides a complete and comprehensive overview of all predictors known in literature. There is a considerable advantage to the novel approach we took in this study to pool odds ratios of predictors. This is especially demonstrated by the pooled results for gender and dorsal comminution. Table 3 shows evidence for both predictors seems inconclusive; however, by pooling we found that both are significantly associated with secondary displacement. Gender is not commonly addressed in popular definitions of an unstable distal radius fracture (Cooney et al., 1979; Lafontaine et al., 1989; Poigenfürst, 1980). Moreover, gender has been refuted as a predictor by several studies (Table 3). It is possible that by pooling, we identified a predictor that did not previously reach significance due to small study sizes. Nevertheless, the effect of female gender could be mitigated when accounting for age, because women reach a higher age than men and are more prone to suffer from osteoporosis. Thus the association between gender and fracture instability is probably indirect and should not be interpreted as direct causality.

This study has several limitations. The majority of studies that we included focused primarily on secondary displacement. Consequently, these studies only described patients who were treated conservatively. Patients treated initially with an operation were probably not included. This last group is likely to include the most unstable fractures. In our opinion, this limitation mostly applies to studies performed after the introduction of volar locking plates in 2000. Conversely, some patients might have been treated operatively who would have achieved excellent results with conservative treatment. Unfortunately, the decisions to perform primary surgical fixation and exclude these patients were only reported in a few studies. Intra-articular involvement, volar fracture displacement or open fractures were most commonly mentioned.

A limitation regarding the meta-analysis is the variation among study populations: four studies provided one common odds ratio for both non-displaced and displaced fractures; one study only provided odds ratios for displaced fractures; and one study reported separate odds ratios for displaced and minimally displaced fractures. We were unable to extract separate odds ratios for non-displaced and displaced fractures and therefore combined these data.

Of particular note is the definition of secondary displacement varied considerably across studies (Table 2). All studies included an alteration in dorsal angulation as a criterion of displacement in their definition. However, cut-off values varied from an absolute dorsal angulation of 10° or 15° from neutral to a change of 5°. The pooled odds ratios we found should therefore be interpreted with some caution.

Despite the obvious variability among studies, the statistical heterogeneity tests showed an I² of 48% or less for all analyses but one. The heterogeneity for dorsal comminution was considerable (I² = 78%). A possible explanation for this could be that the assessment of the presence of dorsal comminution was more subjective and variable across studies than other predictors, such as age and gender.

These limitations emphasize the need for consistency of definitions, measurement methods and a structured follow-up for patients with distal radius fractures. Standardization would allow easier comparison of studies and contribute to a higher level of evidence.

This systematic review provides a comprehensive overview of all known predictors and non-predictors of secondary displacement in patients with distal radius fractures. We have demonstrated that, despite their popularity as predictors of secondary displacement, distal radius fractures with an associated ulna fracture, a dorsal angulation >15° or >20° from neutral and intra-articular fractures do not have an increased risk of secondary displacement. We did find a significantly increased risk of secondary displacement for patients older than 60–65, women and fractures with dorsal comminution. Our results can provide a good basis for surgeons to inform patients on the probability of secondary displacement and therefore the chances of success from conservative treatment. This will facilitate shared decision making between patient and surgeon. Nevertheless, since secondary displacement does not always entail poor functional results after conservative treatment, especially in elderly patients (Bartl et al., 2014), future studies should focus on pooling important predictors of functional outcome.

Footnotes

Supplementary material

The online appendices are available at .

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Abbaszadegan

Jonsson

Sivers

KV.

Prediction of instability of Colles’ fractures. Acta Orthop Scand. 1989, 60: 646–50.

Adolphson

Abbaszadegan

Jonsson

Computer-assisted prediction of the instability of Colles’ fractures. Int Orthop. 1993, 17: 13–5.

Alemdaroglu

Iltar

Aydogan

Say

Kilinc

Tiftikci

Three-point index in predicting redisplacement of extra-articular distal radial fractures in adults. Injury. 2010, 41: 197–203.

Altissimi

Mancini

Azzara

Ciaffoloni

Early and late displacement of fractures of the distal radius: the prediction of instability. Int Orthop. 1994, 18: 61–5.

American Academy of Orthopaedic Surgeons Board of Directors, 5 December 2009. The treatment of distal radius fractures, guideline and evidence report, 2014.

Arora

Lutz

Deml

Krappinger

Haug

Gabl

A prospective randomized trial comparing nonoperative treatment with volar locking plate fixation for displaced and unstable distal radial fractures in patients sixty-five years of age and older. Orthop. 2012, 35: 50–1.

Bartl

Stengel

Bruckner

Gebhard

, ORCHID Study Group. The treatment of displaced intra-articular distal radius fractures in elderly patients. Dtsch Arztebl Int. 2014, 111: 779–87.

Benoist

Freeland

AE.

The shelf sign indicating instability in minimally displaced extraarticular distal radial fractures. Orthop. 1995, 18: 1125–6.

Beumer

McQueen

MM.

Fractures of the distal radius in low-demand elderly patients: closed reduction of no value in 53 of 60 wrists. Acta Orthop Scand. 2003, 74: 98–100.

10.

Blakeney

Webber

Emergency department management of Colles-type fractures: a prospective cohort study. EMA. 2009, 21: 298–303.

11.

Camelot

Ramare

Lemoine

Saillant

Orthopedic treatment of fractures of the lower extremity of the radius by the Judet technique. Anatomic results in function of the type of lesion: apropos of 280 cases. Rev Chir Orthop Reparatrice Appar Mot. 1998, 84: 124–35.

12.

Clayton

Gaston

Ralston

Court-Brown

McQueen

. Association between decreased bone mineral density and severity of distal radial fractures. J Bone Joint Surg Am. 2009, 91: 613–9.

13.

Cooney

3rd Linscheid

Dobyns

JH.

External pin fixation for unstable Colles’ fractures. J Bone Joint Surg Am. 1979, 61: 840–5.

14.

Earnshaw

Aladin

Surendran

Moran

. Closed reduction of colles fractures: comparison of manual manipulation and finger-trap traction: a prospective, randomized study. J Bone Joint Surg Am. 2002, 84-A: 354–8.

15.

Einsiedel

Freund

Sander

Trnavac

Gebhard

Kramer

Can the displacement of a conservatively treated distal radius fracture be predicted at the beginning of treatment. Int Orthop. 2009, 33: 795–800.

16.

Fenyo

Johansson

Secondary displacement of reduced distal radius fractures. Acta Orthop Scand. 1974, 45: 76–81.

17.

Hayden

van der Windt

Cartwright

Cote

Bombardier

Assessing bias in studies of prognostic factors. Ann Intern Med. 2013, 158: 280–6.

18.

Higgins

JPT

Green

. Cochrane handbook for systematic reviews of interventions Version 5.1.0 [updated March 2011], The Cochrane Collaboration, 2011.

19.

Hove

Solheim

Skjeie

Sorensen

FK.

Prediction of secondary displacement in Colles’ fracture. J Hand Surg (Edinburgh, Scotland). 1994, 19: 731–6.

20.

Huguet

Hayden

Stinson

. Judging the quality of evidence in reviews of prognostic factor research: adapting the GRADE framework. Syst Rev. 2013, 2: 71,4053–2–71.

21.

Jenkins

NH.

The unstable Colles’ fracture. J Hand Surg Br. 1989, 14: 149–54.

22.

Kulej

Dragan

. Efficacy of closed reduction and maintenance of surgical outcome in plaster cast immobilization in different types of distal radius fractures. Orthop Traumat Rehabil. 2007, 9: 577–90.

23.

Kumar

Penematsa

Sadri

Deshmukh

SC.

How many clinic visits does it take to treat distal radial fractures?

Int Orthop. 2008, 32: 91–6.

24.

Lafontaine

Hardy

Delince

Stability assessment of distal radius fractures. Injury. 1989, 20: 208–10.

25.

LaMartina

Jawa

Stucken

Merlin

Tornetta

3rd . Predicting alignment after closed reduction and casting of distal radius fractures. J Hand Surg Am. 2015, 40: 934–9.

26.

Leone

Bhandari

Adili

McKenzie

Moro

Dunlop

RB.

Predictors of early and late instability following conservative treatment of extra-articular distal radius fractures. Arch Orthop Trauma Surg. 2004, 124: 38–41.

27.

Mackenney

McQueen

Elton

Prediction of instabilityin distal radial fractures. J Bone Joint Surg Am. 2006, 88: 1944–51.

28.

Makhni

Ewald

Kelly

Day

CS.

Effect of patient age on the radiographic outcomes of distal radius fractures subject to nonoperative treatment. J Hand Surg. 2008, 33: 1301–8.

29.

Moher

Liberati

Tetzlaff

Altman

, PRISMA Group. Preferred reporting items for systematic reviews andmeta-analyses: the PRISMA statement. Int J Surg. 2010, 8: 336–41.

30.

Myderrizi

Factors predicting late collapse of distal radius fractures. Malays Orthop J. 2011, 5: 3–7.

31.

Nesbitt

Failla

Les

Assessment of instability factors in adult distal radius fractures. J Hand Surg. 2004, 29: 1128–38.

32.

Oskarsson

Aaser

Hjall

Do we underestimate the predictive value of the ulnar styloid affection in Colles fractures?

Arch Orthop Trauma Surg. 1997, 116: 341–4.

33.

Phillips

Al-Shawi

Restoration of the volar cortex: predicting instability after manipulation of distal radial fractures. Injury. 2014, 45: 1896–9.

34.

Plaweski

Lantuejoul

Verjux

. Articular fractures of the distal radius: results according to anatomy and treatment. Main. 1997, 2: 11–8.

35.

Poigenfuerst

Tuchmann

The significance of injuries to the ulnar ligaments of the wrist in Colles’ fractures. Handchirurgie. 1978, 10: 121–5.

36.

Poigenfürst

Brüche am distalen Unterarmende. Einteilung der Bruchformen und Indikation. Hefte Unfallheilkd. 1980, 148: 53–9.

37.

Porter

Stockley

Fractures of the distal radius. Intermediate and end results in relation to radiologic parameters. Clin Orthop Relat Res. 1987, 220: 241–52.

38.

Robin

Ellington

Jupiter

Brennan

ML.

Relationship of bone mineral density of spine and femoral neck to distal radius fracture stability in patients over 65. J Hand Surg Am. 2014, 39: 861,6.e3.

39.

Tahririan

Javdan

Nouraei

Dehghani

Evaluation of instability factors in distal radius fractures. J Res Med Sci. 2013, 18: 892–6.

40.

Wadsten

Sayed-Noor

Englund

Buttazzoni

Sjoden

. Cortical comminution in distal radial fractures can predict the radiological outcome: a cohort multicentre study. Bone Joint J. 2014, 96-B: 978–83.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.00 MB