Introduction

Many factors influence the clinical result following a distal radius fracture such as gender, age, injury compensation, education, other medical comorbidities, and the characteristic of injury. As such, much of the literature examining distal radius fractures (DRF) focuses on predicting which factors are associated with successful patient outcomes.

Traditionally, fracture severity and anatomic reduction have been measured using planar radiographs. Gartland and Werley et al. published a landmark paper in 1951 which investigated radiographic measures following distal radius fractures [15]. They used a set of radiographic measures that are typically altered in a distal radius fracture injury. Many subsequent studies have examined the association of pre-reduction and post-reduction radiographic measures with functional or patient-reported outcomes. A previous study examining the relative risk of a poor outcome following mal-alignment of the distal radius concluded that patients at all ages had a higher risk of a poor outcome with mal-alignment when compared with those with acceptable alignment [18]. However, controversy exists in the literature surrounding the impact of radiographic measures on predictors of patient outcomes, particularly in the older population. Grewal et al. reported that the relative risk of poor outcomes with mal-alignment decreased with increasing age but was not statistically significant in patients older than 65 years [18]. Similarly, Anzarut et al. in a cohort of patients older than 50 years reported that acceptable radiographic reduction was not associated with better generic physical or mental health status, lower disability, or greater satisfaction [4]. Other studies suggest that substantial deformity in the alignment of the distal radius is needed before significant alteration in wrist function is evident [15].

The discrepancy between “form and function” has been attributed to a lack of highly powered studies employing standardized outcome measures while controlling for confounding variables such as fracture type and age [18, 19] or lack of standardized radiographic measures [27, 32]. Previous studies have attempted to establish normative guidelines for radiographic measurements following distal radius fractures [27, 32, 57]. However, it is unclear if standardized approaches for taking radiographic measures are consistently being used or even if there is consistent interpretation of these radiographic measures.

The objective of this structured literature review was to determine how radiographic measures are used in evaluating patients with distal radius fractures. Additionally, we sought to assess the variability in the radiographic measurement properties. Specifically, the objective was to examine the extent and range of radiographic parameters measured in research studies investigating distal radius fractures as well as their acceptability criteria and measurement properties using a systematic literature search and structured data extraction process.

Material and Methods

Literature Search and Study Identification

A literature search was conducted using the Evidence-based Medicine Reviews (Embase) and PubMed with publication dates up to and including December 2013. The search was limited to full-text-only publications, written in English. The following keywords were used to search all databases for eligible studies: Distal Radius Fracture OR Wrist Fracture AND Radiographic Outcome OR x-ray OR Imaging OR Measure OR Mal-Alignment Or Alignment. The first stage of study identification was reviewing the titles listed from Embase and PubMed using the specified keywords. In total 988 titles were reviewed (Fig. 1). Studies were excluded if they involved nonhumans, if they included less than 100 participants, if they were review articles, and if there were published in conference proceedings or as a thesis or dissertation. Additionally, studies were excluded if they did not include radiographic measurements of the distal radius, investigated medical imaging technologies other than X-ray or were surgical technique papers.

Fig. 1
figure 1

Structured review flowchart

Full size image

Study Selection

In total, 263 studies were included after removing two duplicate studies listed in both Embase and PubMed. All 263 abstracts were collected and screened. After abstract screening, 209 studies were excluded and 54 full-text articles were then considered eligible for data extraction. Twenty-three studies were then further excluded if after reading the full article (resulting in a final of 31 studies included), it was evident that they did not meet the inclusion criteria (16 had a sample size less than 100, 5 articles did not have actual radiographic measures, 1 was not in English, and 1 was a description of a surgical technique) (see Fig. 1).

Data Collection Process

The data extraction and review process was conducted using a standardized data extraction form that was developed for this review (Table S1). All articles were reviewed by the first author. Where there was uncertainty in the extraction, the senior author was consulted.

Results

Study Demographics

In total, 31 studies were included in the data extraction process. The mode of the sample size of the studies included in this review was 124 participants. Of the studies not included (231), the mode of the sample size was 20 participants and the overall average was 38 participants. Additionally, 80 % of articles not included in the study had 50 or less participants included. Table 1 lists the authors, titles, and journal reference in alphabetical order. The majority of the articles reviewed investigated adult patients (19–64 years) with distal radius fractures (84 %) (3/31 studies examined older patients ≥65, 2/31 studies investigated pediatric patients (≤18 years), and were conducted in Europe (15 studies), North America (11 studies), and Asia (5 studies)). The majority of participants/patients in these studies were woman (70 % woman). Detailed information describing the study location, sample size, age, and gender is shown in Table 2. The frequency of studies investigating radiographic measures of distal radius fracture alignment increased from 2007 to 2009 (Fig. 2) and then decreased from 2009 to 2013.

Table 1 Summary of Studies Investigating Radiographic Measures and Distal Radiographic Fractures
Full size table
Table 2 Study Demographics
Full size table
Fig. 2
figure 2

Number of studies by year

Full size image

How Are Radiographic Measures Used in Studies Evaluating Patients with DRFs?

Table 2 lists the designation of each study design as a (radiographic) technique paper, prognostic or explanatory paper (for definitions of these see Table S1). Fifty-eight percent (58 %) of the studies had prognostic analysis present in their study design and reported results. The majority of these prognostic studies compared radiographic measures (typically at baseline, and this would also include radiographic fracture classification at baseline) to radiographic measures at a follow-up visit (44 % of all prognostic studies). Radiographic measures (also measured at baseline, or fracture classification measured at baseline) were also compared to patient outcome measures at follow-up (33 % of all prognostic studies). Fracture classification or baseline radiographic measures (usually of alignment, reduction or mal-alignment, or loss of reduction) were used to predict treatment requirements or characteristics (23 %). Thirty-seven percent of all the studies had explanatory (cross-sectional design). Similar to the prognostic studies, the majority of these studies compared radiographic measures to other radiographic measures (46 %). This explanatory analysis typically compared fracture classification to radiographic measures of mal-alignment. Radiographic measures were also compared cross-sectionally to patient outcomes and to treatment requirements or characteristics in 30 and 24 % of the explanatory studies, respectively.

What Is the Variability in the Extent, Range, and Acceptability Criteria of Radiographic Measures Examined in Studies Evaluating Patients with DRFs?

Radiographic Parameters

The distribution of radiographic measures examined in these studies is shown in Fig. 3. Overall, there is an even distribution in the type of radiographic measures with the most common being dorsal/volar angulation (22 studies) and radial inclination (19 studies). Table 3 lists the radiographic measures used in each study. When considering the country of origin of each study, there was no consensus with regard to the radiographic measures employed. Only one study examined a composite radiographic measure that combined two or more of these radiographic measures and called this the “absolute degree of deviation” which combined radial inclination and volar tilt [47]. Radial inclination was also termed radial deviation and radial angle but was always measured in the posterior-anterior radiograph. Volar and dorsal angulation was always measured simultaneously as it described the angulation of the radial articular surface in the sagittal plane and was examined in the lateral radiographs. Volar tilt was also called palmar tilt or these terms were combined and were called dorso-volar plane angulation. Intra-articular step and gap were alternatively called articular congruency and were only applicable in intra-articular distal radius fractures. Over half of the studies (18/31) measured additional radiographic parameters as shown in Table 3 and consisted primarily of identification of comminution present in the metaphysis (four studies), measures of carpal mal-alignment (four studies), and inter-carpal angles (three studies).

Fig. 3
figure 3

Frequency of radiographic measures examined

Full size image
Table 3 Radiographic measure
Full size table

Definitions for the established radiographic acceptability were extracted for each study and are also shown in Table 3. Acceptability values for each radiographic measure, when indicated, were used to define mal-alignment, displacement, or mal-union or used to establish successful and unsuccessful radiographic outcomes. Radial height had two acceptable radiographic criteria but had overlap in their range 14 mm vs. 9 (8–14 mm). Radial shortening had acceptability criteria and unacceptability criteria in the range of 2 mm (age of participants in study, 16–60 years) [29, 42] to 5 mm (age of participants in study, 20–89 years) [35, 36, 51] with increased shortening considered unacceptable. Radial inclination had the most unique definitions of acceptability and unacceptability criteria, but these definitions were similar in magnitude. Ulnar variance had two definitions for acceptability and indicated that less ulnar variance was best (aligned with the distal radius with neutral or slightly negative ulnar variance was best). Volar and dorsal angulation were measured simultaneously with a general consensus that any deviation from the native volar tilt into dorsal tilt was considered unacceptable with increasing dorsal angulation representing the largest deviation from the native volar tilt. The presence of intra-articular step and intra-articular gap less than 2 mm was considered acceptable in all studies and deformities larger than this were considered to be unacceptable radiographically.

The terms “radial height/shortening/length and ulnar variance” were not consistently employed or described. Table 4 lists the referenced techniques and the description of the referenced techniques related to radial shortening/height or ulnar variance cited in the studies included in this review. Ulnar variance as described in this review (shown in Table 3) study is measured on a posterior-anterior radiograph and measures the distance between horizontal lines (that are perpendicular to the long axis of the radius/ulna) drawn from the distal ulnar and radial articular surfaces (at the level of the distal radioulnar joint). After a detailed examination of the referenced techniques cited in this review, it appeared as though four studies measured ulnar variance according to this definition but instead called it radioulnar index [3] or radial shortening [12, 14, 33].

Table 4 Ulna Variance, Radial Shortening and Radial Length Referenced Techniques
Full size table

Radiographic Measurement Characteristics

Table 5 lists the characteristics of the radiographs (when measured, type of projection and the use of bilateral comparison) as well as the reliability and measurement properties of the radiographic measures. Lateral and posterior-anterior radiographs were used in 17 of the articles (55 %) whereas lateral, posterior-anterior, and oblique radiographs were used in four studies (13 %). Ten articles (32 %) did not indicate the type of projection used. Twenty-three percent of the time bilateral radiographs were acquired of the noninjured wrist, and this comparison was typically made at baseline to compare radiographic measures of radial height for example or during the last follow-up visit. Twenty-one (68 %) of all the articles specified who was taking the radiographic measures (the other ten articles did not specify). Five of these articles had radiologists take the radiographic measures. The majority of the studies had either co-authors (nine), research nurses (one), had an independent assessor (one), research assistant (two), or the attending surgeon/resident/fellow (five) take the radiographic measures.

Table 5 Radiographic Measures Characteristics
Full size table

Measures of reliability were assessed in four studies and three additional studies referenced the reliability reported in previous studies. This reliability component typically involved reviewing a subset of the radiographic measures and comparing retest measures between observers or within the same observer (approximately half of the articles either referenced or published a standardized technique to specific radiographic measures (13 studies, 52 %) (ICC 0.70–0.99). Table 6 lists the articles that either had detailed description (drawings or text) (11 studies, 35 %) or that referenced standardized techniques for each of the radiographic measures (13 studies, 42 %). Six studies used Kreder et al. 1993 for radial height, ulnar variance, and dorsal and volar tilt [27]. Table 6 also lists the other specific reference standards for acceptability in radiographic measures and if this decision was based on a literature review or clinical experience.

Table 6 Referenced radiographic measurement techniques and acceptability criteria
Full size table

Table 7 lists the type of fracture classification system and degenerative scale used in each study. Twenty-three of the articles (74 %) classified the fractures with the majority of the studies using the AO classification system (19 articles). All studies that used a degenerative scale used the one developed by Knirk and Jupiter (26) (6 studies).

Table 7 Fracture Classification and Degenerative Scales Employed
Full size table

Discussion

The results of this structured review indicate that there appears to be a commonly used set of radiographic parameters that are routinely measured. The primary measures described in all studies are those translations or angulations that are most consistently found following a dorsally angulated distal radius fracture (Colles’ fracture) and are also those parameters that were described in the initial Gartland and Werley study. However, despite having a commonly used set of radiographic measures, approximately 65 years later, no consensus has been established as to the relative impact radiographic measurements have on wrist function, or standardization of measurement technique or acceptability criteria. Even after focusing on large clinical studies, we found substantial variation in how radiographic images are being measured and interpreted.

When examining the study demographics of the articles included in this review, the majority of these patients/participants examined were adult woman which is consistent with the cumulative incidence of 15 % in women and 2.5 % for men [1, 10, 46, 49, 55]. Also, the number of studies per year taking radiographic measurements in cohorts of 100 or more people was expected to increase with time due to the advances in accessibility of digital radiographs. However, the results of this review did not indicate this increased trend.

After examining the terminology used in the radiographic measures employed in the studies included in this review, it appears that there is some discrepancy in the measurement of ulnar variance and radial shortening. Ulnar variance is described as the measurement between the two horizontal lines visible on the posterior-anterior radiograph (and perpendicular to the long axis of the radius/ulna) between mid-articular surface of the distal radius and the distal articular surface of the ulna. The exact same definition of this radiographic measure was also given to definition of radial shortening [12, 14, 33] or radioulnar index [3]. Radial shortening was also described in the studies included in this review as the distance between two horizontal planes (posterior-anterior radiograph) from the radial styloid to the distal ulnar articular surface. Warwick et al. compared three techniques used to measure radial shortening as described by Stewart et al. [2, 52] (distal ulnar styloid to distal radial styloid), by Gartland and Werley [15] (radial styloid to distal ulnar surface), and by Altissimi [3] (although termed “radioulnar index”) (at the level of the distal radioulnar joint) and determined that the definition of radial shortening which is measured at the level of the distal radioulnar joint [3] should be adopted as this was the measurement technique that was most correlated with measured clinical outcomes. However, based on the results of this structured review, it appears that confusion surrounding the exact description of ulnar variance and radial shortening persists. We therefore suggest standardized, measurement techniques as shown in Table 3 for radial shortening, ulnar variance, radial inclination, and volar/dorsal tilt.

Kreder et al. and MacDermid et al. recognized the importance of standardization of radiographic measurements for use in studies relating functional outcome to radiographic alignment and also noted that despite the pervasiveness of this imaging technique and measures in patients with distal radius fractures, little work has been published that comments specifically on the reliability and standardization of the these radiographic measures [27, 32]. Furthermore, the use of acceptability criteria or the development of benchmark values which dictate the efficacy of treatment options seems unwarranted when the reliability of these measurements is not reported is unknown or the measurement technique is not standardized. Based on the results of this review, three studies investigated the reliability of the radiographic measures and three additional studies referenced previous studies which investigated the reliability of the measurements examined.

A limitation of our review was that we decided to limit the inclusion of studies to those studies which examined 100 or more participants. The purpose of this review was to examine the extent and range of radiographic parameter used in studies investigating patients with distal radius fractures. Based on the results of our initial literature search, it appeared that the majority of studies included sample sizes less than 50 participants. We decided to limit the inclusion criteria for this structured review to those studies that had 100 more participants included as we assumed that larger high-quality studies would be employing radiographic measures that were representative of the types and uses of radiographic measures employed in studies with smaller sample sizes

Anatomic alignment and radiographic measures of restoration of joint alignment has been the focus of much of the research surrounding fractures of the distal radius. Restoration of the alignment of the joint continues to be a primary goal in treating patients with distal radius fractures. However, definitions of acceptability and benchmark indicators for good patient outcomes vary greatly as well as definitions of mal-union, re-displacement, and loss of reduction. Since many studies evaluated radiographic measures over time to measure or monitor treatment response, and since multiple raters are often involved, the lack of standardization of methods is a substantial barrier to defining prognosis and outcomes in patients with distal radius fractures.