Abstract
Aims
To evaluate the diagnostic characteristics and reliability of radiological methods used to assess scaphoid fracture union through a systematic review and meta-analysis.
Methods
MEDLINE, Embase, and the Cochrane Library were searched from inception to June 2022. Any study reporting data on the diagnostic characteristics and/or the reliability of radiological methods assessing scaphoid union was included. Data were extracted and checked for accuracy and completeness by pairs of reviewers. Methodological quality was assessed using the QUADAS-2 tool.
Results
A total of 13 studies were included, which were three assessed radiographs alone, six CT alone, and four radiographs + CT. Diagnostic sensitivity was assessed by CT in three studies (0.78, 0.78, and 0.73) and by radiographs in two studies (0.65, 0.75). Diagnostic specificity was assessed by CT in three studies (0.96, 0.8, 0.4) and by radiographs in two studies (0.67, 0.4). Interobserver reliability was assessed for radiographs by seven studies (two fair, four moderate, and one substantial) and for CT in nine studies (one fair, one moderate, six substantial, and one almost perfect).
Conclusion
There is evidence to support both the use of CT and radiographs in assessing scaphoid fracture union. Although CT appears superior in terms of both its diagnostic characteristics and reliability, further research is necessary to better define the optimal clinical pathways for patients.
Cite this article: Bone Jt Open 2025;6(3):246–253.
Take home message
There is evidence to support both the use of CT and radiographs in the diagnosis of scaphoid fracture union.
CT has better diagnostic characteristics and reliability than radiographs.
Future studies should focus on how best to use CT to improve current pathways of care.
Introduction
Scaphoid fractures are the commonest carpal fracture1 and account for approximately 15% of all acute wrist injuries.2,3 The negative consequences of nonunion for patients and healthcare providers are significant.4 Consequently, it is important to be able to diagnose union and nonunion both accurately and reliably. The early diagnosis of union can help to enable early mobilization, reassure patients, and save money for healthcare providers by reducing the need for costly further clinical attendances,5 while delays in diagnosing nonunion can reduce the chances of surgery achieving union and result in further degenerative change occurring adjacent to the scaphoid.4,6 The radiological assessment of union continues to pose challenges, due to its unique size, shape, and orientation. To our knowledge, no systematic review has previously attempted to summarize the evidence in this area. The aim of this systematic review of the literature is to answer the following questions relating to the different radiological methods used to assess scaphoid fracture union: 1) what are the diagnostic characteristics of these methods?; and 2) what is the reliability of these methods?
Methods
This systematic review is reported in accordance with the PRISMA statement (Supplementary Table iii).7 The protocol was developed prospectively, and peer reviewed locally before registration on the PROSPERO database (CRD42022341571).
Data sources and searches
We used the Ovid interface to search two key bibliographic biomedical databases. Searches were run on MEDLINE (1946 to present) and Embase (1974 to present). No date or language limits were applied. We also searched the Cochrane Register of Controlled Trials and the Cochrane Database of Systematic Reviews. The searches were run on 14 June 2022. Sets of synonyms were produced in line with the distinct elements of our research question. There are three major aspects. We therefore selected thesaurus terms and free-text terms for each. Scaphoid was the first, fracture the second, and union or nonunion was the third concept. Synonyms for scaphoid included the MeSH terms scaphoid bone and scaphoid fracture. Synonyms for fracture included break or broken. Synonyms for union or nonunion included heal and malunion.
Inclusion/exclusion criteria
The aim was to include any study reporting on radiological methods assessing scaphoid fracture union. The diagnostic test could include any form of radiological investigation. In terms of the association between diagnosis of union and outcome, only studies that described the diagnostic attributes (such as sensitivity, specificity, accuracy) and/or the reliability of the radiological assessment were included. The search aimed to identify any study relating to scaphoid fractures in adults (aged ≥ 18 years). Articles with the following criteria were excluded: review articles, duplicate results, lack of full access to the original article, articles not available in English, studies not published as a full article such as conference abstracts and letters, and case studies.
Selection of studies
Duplicates were removed and relevant studies identified from the search were imported into Covidence for screening.8 Studies were independently screened by title and abstract by two authors (WT, GS). The references of all included studies and all relevant review articles on the topic were also reviewed to identify other potential studies for inclusion. This was followed by a full-text evaluation of the selected studies from the first selection step by these authors. Disagreement between the two reviewers was solved by consensus involving a third author (BD).
Data extraction
Two reviewers independently extracted data (WT, GS). Data were extracted using a custom data extraction sheet in Covidence. The custom data extraction sheet was specifically designed to extract data relating to study design, participant characteristics, clinical setting, study design, target condition definition, index test, reference standard, and sample size. If interobserver reliability had been analyzed, Kappa scores were noted. Intraobserver reliability was also reported, if performed. Values ≤ 0 were considered to indicate no agreement, 0.01 to 0.20 none to slight, 0.21 to 0.40 fair, 0.41 to 0.60 moderate, 0.61 to 0.80 substantial, and 0.81 to 1.00 almost perfect agreement. Rossi et al9 reported correlation between observers (Spearman rank). Furthermore, 2 × 2 data (true positive, false positive, false negative, true negative) with estimates of diagnostic accuracy (sensitivity and specificity) of fracture union and CIs were recorded. Where 2 × 2 data were not provided, it was calculated where sufficient data were provided to enable this. Any inconsistencies between the two reviewers’ forms were resolved by consensus discussion. A third reviewer was available for any disagreement that could not be resolved by this initial discussion, but this was not needed.
Risk of bias/quality assessment
Included studies were assessed for risk of bias/quality by two independent raters (GS, WT) using the QUADAS-2 tool.10 All domains were scored for the studies that had assessed diagnostic characteristics; however, certain domains were not applicable to the studies that had solely assessed reliability.
Statistical analysis
Descriptive analysis was performed for all data relating to the test diagnostic characteristics and reliability to facilitate narrative interpretation and comparison across studies. It was only possible to obtain 2 × 2 data for three of the four studies that assessed diagnostic characteristics. A study by Farracho et al5 did not provide sufficient data to enable this, and the corresponding author did not respond to our request for further data. As this left one study with 2 × 2 for radiographs and two studies with 2 × 2 data for CT,11-13 a decision was made that pooling the data for meta-analysis would not be of additional value.
Results
Study selection
This process is shown by Figure 1. In total, 4,297 studies were found based on our search term, including 154 duplicates. Two authors reviewed all the remaining 4,143 abstracts. Of these, 4,040 studies were subsequently excluded as these did not meet the inclusion criteria, leaving 97 articles for final review. Articles were excluded for the following reasons: two as the full text was not accessible, two had the incorrect patient population, four were conference abstracts, ten were not written in English, and 32 were the incorrect study design or outcome. This left 43 articles, of which only 13 had objective data on the diagnostic characteristics and/or reliability of radiological assessment of scaphoid union for the systematic review.
Fig. 1
Study selection process.
Study characteristics
Study details, participants, and interventions are described in Table I. The radiological methods used to assess scaphoid union from the 13 studies included radiographs and CT. Conservative treatment in cast with or without thumb immobilization was used in part or as the sole treatment of patients in 11 studies. Treatment also included headless compression screw fixation for the primary treatment of acute fracture14,15 and in patients for the treatment of nonunion,15 Russe bone grafting for scaphoid nonunion cases9, and pulsed electromagnetic stimulation in addition to cast immobilization.16 The details regarding the radiological assessment of union methods (where provided), the number and details of observers, the details of the union assessment, and reference standard for diagnostic studies are shown in Supplementary Table ii.
Table I.
Study details including patients and interventions.
| Author, yr | Participants, n |
Inclusion criteria | Scaphoid fracture location | Interventions | Comparators | Outcomes | Timepoints |
|---|---|---|---|---|---|---|---|
| Buijze et al11 | 30 | Scaphoid waist fractures (20 known to be united at six months following injury) | Scaphoid waist fractures | Conservatively treated up to six weeks after injury (all 30 patients) | Ten patients had subsequent operative fixation for operatively confirmed nonunion | United or ununited | CT scan performed minimum six weeks after injury |
| Dias et al17 | 20 | Scaphoid waist fractures | Waist fractures | Conservatively treated | Nil | Union and crossing trabeculae | 12 weeks post-injury radiography |
| Dias et al15 | 439 | Bicortical fracture of the scaphoid waist on scaphoid radiography < two weeks of presentation July 2013 to September 2017 |
Bicortical waist of scaphoid fracture < 2-mm step | Percutaneous or ORIF with headless compression screw | Conservative treatment in cast (with or without thumb) | Radiography: United, Almost Partial Probably not Not united CT: United Partial union Not united |
Six and 12 weeks Final imaging at 52 weeks (union assessment) |
| Drijkoningen et al 18 | 13 | Treated nonoperatively for nondisplaced scaphoid waist fracture | Nondisplaced scaphoid waist fracture | Conservative treatment with cast | Nil | Percentage of bony bridging | Ten to 12 weeks after injury |
| Farracho et al5 | 52 | Scaphoid fracture diagnosis April 2018 to March 2019 | Scaphoid fracture | Conservative treatment with cast for six weeks (without thumb) | Nil | Union Nonunion |
Six weeks cast then radiography and CT scan, eight weeks clinical review |
| Geoghegan et al19 | 57 | Scaphoid waist fractures visible on radiograph September 2002 to December 2003 | Scaphoid waist fracture displaced (17) or undisplaced (43) | Conservative treatment in cast for four weeks (thumb not included) | Nil | United Ununited | Four weeks for CT scan (further scans at eight, 12, and 26 weeks) |
| Grewal et al20 | 50 | Random sampling of acute scaphoid fractures from radiology tertiary centre database 2004 to 2010 | Nonoperatively treated scaphoid fractures | Conservative treatment in cast | Nil | United (75% to 100%) Partially united (50% to 75%) Tenuously united (≤ 50%) |
CT scan one to 197 days after injury |
| Hannemann et al12 | 44 | Unilateral waist of scaphoid fracture November 2009 to February 2012 |
Scaphoid waist fracture undisplaced or minimally displaced | Conservatively in short arm cast including the thumb | Nil | Amount of consolidation21 Union Partial union Nonunion and Union Nonunion |
CT six to 24 weeks after injury |
| Hannemann et al13 | 47 | May 2006 to March 2008 unilateral undisplaced waist scaphoid fractures < five days of injury via radiographs or CT |
Waist scaphoid fractures | Conservative treatment in cast including thumb | Nil | Union Partial union Nonunion |
Six weeks after injury |
| Hannemann et al16 | 102 | January 2010 to December 2011 undisplaced unilateral scaphoid fracture diagnosed < five days of injury via radiography or CT | Herbert A1, A2, B1, and B2 fractures | Conservative treatment in cast with thumb included plus electromagnetic stimulation | Conservative treatment in cast including thumb and placebo | Union Partial union Nonunion |
Six, nine, 12, 24, and 52 weeks after diagnosis of the fracture |
| Matzon et al14 | 32 | Scaphoid fracture treated with ORIF or scaphoid nonunion treated with ORIF 2012 to 2018 | Proximal, waist and distal pole scaphoid fracture undergoing ORIF | Cannulated headless compression screw fixation | Nil | Healed Partially healed Not healed |
Radiography 3.2 months postoperatively (1.3 to 19.5) and CT scan performed 4.5 days after this (1 to 52) |
| Rossi et al9 | 50 | Scaphoid nonunion treated with Russe bone graft 1977 to 1993 | Scaphoid waist fracture with nonunion | Russe bone grafts (no fixation) With cast immobilization (long arm then short arm with thumb included for up to 20 weeks) |
Nil | Fracture line bridging and degree of lucency both graded 0 to 2 | Radiography 2, 3, 4, 5 to 6, 7 to 12, and 13 to 36 months after surgery |
| Singh et al21 | 100 | Scaphoid fractures visible on initial radiography, October 2001 to October 2003 | Waist and proximal pole scaphoid fractures | Treated nonoperatively in cast excluding thumb for eight weeks ± four weeks if clinical or radiograph concerns | Nil | United Partially united Nonunion |
CT scan at 12 and 18 weeks and if concerns CT after further three months |
-
ORIF, open reduction and internal fixation.
Results of included studies
The test diagnostic characteristics (including accuracy, sensitivity, specificity, and positive and negative predictive values) and reliability (inter- and intraobserver) are described in Table II and Supplementary Table i. Four studies included an assessment of test diagnostic characteristics; this included one study involving just radiographs, one using both radiographs and CT, with the other three studies using solely CT scans.5,11,12
Table II.
Data measurements, characteristics, and reliability.
| Author, yr | Modality | Diagnostic characteristics | Other | Interobserver reliability (Kappa unless otherwise stated) |
|---|---|---|---|---|
| Buijze et al11 | CT | Accuracy 0.84 (95% CI 0.63 to 1.0) Sensitivity 0.78 (95% CI 0.47 to 1.0) Specificity 0.96 (95% CI 0.85 to 1.0) PPV 0.99 (95% CI 0.97 to 1.0) NPV 0.41 (95% CI 0 to 0.84) |
0.66 | |
| Dias et al17 | XR | N/A | Fracture union 0.386 Bridging trabeculae 0.104 |
|
| Dias et al15 | XR + CT | N/A | XR: 0.724 agreement (95% CI 0.673 to 0.775) 0.769 agreement (95% CI 0.721 to 0.818) 0.684 agreement (95% CI 0.631 to 0.737) CT: 0.896 agreement (95% CI 0.861 to 0.94) 0.841 agreement (95% CI 0.799 to 0.882); 0.915 agreement (95% CI 0.883 to 0.947) |
|
| Drijkoningen et al18 | CT | N/A | 0.34 bony bridging 0.31 location bony bridging |
|
| Farracho et al5 | XR + CT | Accuracy XR 0.53 to 0.59; CT 0.61 to 0.63 Sensitivity 0.75 XR and 0.78 CT Specificity 0.4 XR and CT PPV XR PPV = 0.66; CT PPV = 0.67 NPV XR NPV = 0.5; CT NPV = 0.53 |
Radiologist XR = 0.35 Surgeon XR = 0.956 Both XR = 0.543 Radiologists CBCT = 0.803 Surgeons = 0.803 Both CBCT = 0.641 |
|
| Geoghegan et al19 | CT | N/A | 0.77 | |
| Grewal et al20 | CT | N/A | Hand surgeon and fellow 0.62 (95% CI 0.44 to 0.80) Hand surgeon and MSK radiologist = 0.8 (95% CI 0.65 to 0.93) |
|
| Hannemann et al12 | CT | Accuracy N/A Sensitivity 0.73 (95% CI 0.66 to 0.79) Specificity 0.8 (95% CI 0.68 to 0.92) PPV N/A NPV N/A |
Positive likelihood ratio 3.65 Negative likelihood ratio 0.34 Diagnostic odds ratio 10.9 |
Overall = 0.576 (95 % CI 0.399 to 0.753) Union = 0.683 Partial union = 0.502 Nonunion = 0.791 Overall union/nonunion = 0.699 (95 % CI 0.529 to 0.870) Union = 0.793 Nonunion = 0.793 |
| Hannemann et al13 | XR | Accuracy N/A Sensitivity 0.65 (95 % CI 0.54 to 0.75) Specificity 0.67 (95 % CI 0.39 to 0.86) PPV 0.93 (95 % CI 0.83 to 0.97) NPV 0.22 (95 % CI 0.12 to 0.38) |
Overall = 0.583 (95 % CI 0.371 to 0.795) No union = 0.816 (95 % CI 0.321 to 0.999) Union = 0.517 (95 % CI 0.077 to 0.999) Partial union = 0.390 (95 % CI 0.048 to 0.832) |
|
| Hannemann et al16 | CT | N/A | Union = 0.683 (95% CI 0.473 to 0.893) Nonunion = 0.791 (95% CI 0.599 to 0.984) |
|
| Matzon et al14 | XR + CT | N/A | Weighted reliability: XR 0.53 (0.42 to 0.63) XR and CT = 0.59 (0.53 to 0.68) for healed vs partially healed vs not healed |
|
| Rossi et al9 | XR | N/A | The scores given by the two observers were significantly correlated, with correlation values ranging from 0.37 to 0.55 (p < 0.001, Spearman rank correlation) | |
| Singh et al21 | XR + CT | N/A | XR = 0.27 |
-
CBCT, cone beam CT; MSK, musculoskeletal; N/A, not available; NPV, negative predictive value; PPV, positive predictive value.
All 13 studies included an assessment of interobserver agreement. Interobserver agreement was measured using Landis and Koch 1977 Kappa statistics22 if between two observers, and Kappa Fleiss23 if among more than two observers in all but two studies.
Sensitivity of CT scans from six to 24 weeks after injury in assessing union was measured at 0.73 by one study12 and 0.78 by two studies,5,11 which was slightly better than radiographs, where assessments were made six weeks following injury and measured 0.6513 and 0.75.5 The specificity of CT scans varied between 0.45 in a study that assessed scans six weeks following injury, compared with them being much better at 0.8 and 0.9611 in studies that assessed scans from six to 24 weeks after injury; radiographs reported specificity of 0.45 and 0.6713 in studies that only assessed scans at six weeks following injury. Accuracy of CT scan (0.61 to 0.63) was superior to radiographs (0.53 to 0.59) in predicting scaphoid healing at early follow-up (six weeks),5 and was reported as high (0.84) in a further study assessing CT scan only.11
Varying analysis of reliability included: assessment of radiographs or CT scans; assessment of different observers; assessment of differing methods to assess union; and radiographs versus radiographs and CT scans. Only four included an assessment of intraobserver agreement and included results for reading radiographs, CT scans, and a combination of the two.
Overall interobserver reliability was assessed for radiographs by seven studies (two fair, four moderate, and one substantial) and for CT in nine studies (one fair, one moderate, six substantial, and one almost perfect). Dichotomizing results as either union or nonunion, rather than including partial union, improved reliability in two studies with kappa score of substantial: 0.69912 and 0.683.16 Interestingly, in the sole assessment of the quantity of bony bridging on a CT scan, the kappa score was only fair: 0.34.18 Interobserver reliability also improved with grade of assessing observer20 and interpreting both radiographs and CT together. Overall interobserver reliability reported by Dias et al15 was acceptable agreement (mean 72.6%; 95% CI 63.1 to 81.8) with regards to radiographs and good agreement (mean 88.4%; 95% CI 79.9 to 94.7) regarding CT scans performed at 52 weeks. However, there was no separation of results for those patients managed conservatively or with those treated with open reduction and internal fixation.
Risk of bias/quality assessment
The risk of bias and applicability assessment is shown in Table III. The reference standard used in the study by Farracho et al5 was short term and consequently felt to be a reason for concern in terms of both risk of bias and applicability.
Table III.
Quality assessment for each study.
| Risk of bias | Applicability concerns | |||||||
|---|---|---|---|---|---|---|---|---|
| Author, yr | Design | Patient selection | Index test | Reference standard | Flow and timing | Patient selection | Index test | Reference standard |
| Buijze 201211 | Retrospective cohort study | Unclear | Low | Low | Low | Low | Low | Low |
| Dias et al17 | Retrospective randomly selected review | Low | N/A | N/A | N/A | Low | Low | N/A |
| Dias et al15 | Randomized controlled trial | Low | N/A | N/A | N/A | Low | Low | N/A |
| Drijkoningen et al18 | Retrospective review | High | N/A | N/A | N/A | Low | Low | N/A |
| Farracho et al5 | Prospective cohort study | Low | Low | High | Low | Low | Low | High |
| Geoghegan et al19 | Prospective cohort study | Low | N/A | N/A | N/A | Low | Low | N/A |
| Grewal et al20 | Retrospective blinded review | High | N/A | N/A | N/A | Low | Low | N/A |
| Hannemann et al12 | Prospective cohort study | Low | Low | Low | Low | Low | Low | Low |
| Hannemann et al13 | Prospective cohort study | Low | Low | Low | Low | Low | Low | Low |
| Hannemann et al16 | Randomized controlled trial | Low | N/A | N/A | N/A | Low | Low | N/A |
| Matzon et al14 | Retrospective review | High | N/A | N/A | N/A | Low | Low | N/A |
| Rossi et al9 | Retrospective review | High | N/A | N/A | N/A | Low | Low | N/A |
| Singh et al21 | Prospective cohort study | Unclear | N/A | N/A | N/A | Low | Low | N/A |
-
N/A, not available.
Discussion
The key findings of this systematic review are that there is some reasonable evidence to support both the use of radiographs and CT in the diagnosis of scaphoid fracture union, but the diagnostic characteristics and reliability of CT appear superior to that of radiographs. A test’s performance is improved with higher values of both sensitivity and specificity, with a cumulative value of over 1.5 required for it to be deemed useful.24 For CT, this value was reached in all but one study,5 but in neither of the studies that assessed radiographs. The reliability of assessing scaphoid union using CT was also generally superior to that of radiographs, with most studies assessing CT demonstrating substantial reliability versus most demonstrating moderate reliability for radiographs.
It is important that these findings should be considered in the context of the quality and exact nature of imaging obtained. For example, Supplementary Table ii illustrates some of the variability in the nature of the CT scans that were analyzed, with incomplete information provided regarding the type and age of CT scanner used, the thickness and number of the slices obtained, whether the bone window was used (which can help to identify bone edges to calculate proportions of union), and whether multiplanar reconstructions were used. The same is true for plain radiographs, where the number, type, and quality of radiographs assessed showed variation across the relevant papers. Furthermore, the quality of the reference standard used is also clearly of critical importance, with the study by Farracho et al5 being significantly flawed in using a two-month timepoint in this regard. Notably, this study produced values of specificity of 0.4 for both radiographs and CT, results which should not carry significant weight due to the flawed nature of this reference standard. The timing of imaging is also of critical importance, with it being well recognized that the accuracy of diagnosing union will increase as the time to radiological assessment increases. Differences in the time to imaging may explain some of the observed superiority of CT over radiographs; however, it does not account for the superior reliability as this was observed with equivalent time to follow-up.
Currently in the UK, only 28% of centres use CT scans as their first line in the radiological assessment of scaphoid union.25 The challenge of confirming scaphoid union, as well as defining when it is appropriate to discontinue immobilization and commence specific activities, is a very important topic due to the potential negative implications of developing a painful scaphoid nonunion. It is made more complex by the heterogeneity of the populations in this systematic review (acute fractures versus treatment of nonunion, and nonoperative versus various forms of operative treatment). For example, internal fixation changes the biomechanics and subsequent physiology of fracture healing, which will influence the way in which union is assessed and interpreted. The role of clinical examination remains unclear, as there is a paucity of evidence in this domain; however, it was not within the scope of this review to ask this question.
The complexities around this topic continue to be researched and discussed, including cadaveric analysis26 and finite element analysis studies of scaphoid union.27 The British Society of Surgery for the Hand has recently produced standards for the follow-up of confirmed scaphoid fractures.28 In addition, it is important to consider that there are some disadvantages to the routine use of CT over radiographs, which include higher costs, less availability, and greater radiation exposure. It should be noted that the radiation exposure of a modern wrist CT is very low indeed.29 Notably, in the NHS, the estimated cost of a CT is relatively low at GBP £60, which is far lower than the cost of a follow-up clinic appointment.30
Limitations of this review include the absence of data in nine studies relating to sensitivity and specificity and thus, positive and negative predictive values. This review included six retrospective articles, five prospective cohort studies, and only two randomized controlled trials, meaning that almost half were subject to potential spectrum bias as the patients were a known cohort prior to analysis being performed. As stated previously, each study has attempted to answer a different question in a different population, as well as the timeframe from injury or operation to imaging varying hugely, from four weeks19 to 52 weeks.15 This study heterogeneity is worth acknowledging, but in this context, we feel that it is reasonable to ascribe some meaning to the descriptive trends in the data. Certainly, the optimal patient pathway remains unclear given this heterogeneity, making a strong case for better designed prospective research in this area to define the role of CT with more clarity. It is likely that the optimal pathways would not be homogenous, given that factors such as fracture location (proximal vs distal) and mode of treatment (cast vs acute surgery vs delayed surgery) would likely require different approaches.
The definition of scaphoid union, partial union, or nonunion was not standardized across all papers, although Singh’s grading of union was used in several of the studies assessed.13,14,16,18,21 The dichotomy of healed or not healed (which may or may not be argued as helpful) was used in only three studies.5,11,19 Fundamentally, union is a process of bridging trabeculae, consolidation, and subsequent remodelling, but defining what can be measured on imaging and when to allow for return to function is the critical question.
There may be more uncertainty in this area than we imagine, in that although in many cases union reliably progresses once a certain degree of partial union has been reached, it may well be that partial union has the potential to regress in certain circumstances. The observers were mostly experts or individuals specifically trained to interpret such imaging, which potentially makes these results only generalizable to those with similar levels of experience and training. This is reinforced by the results showing an improved interobserver reliability,20 with improved grade of assessor. To combat this issue, educating observers in the technique of measuring the area of fracture and the area of trabecular bridging on appropriate CT slices would be a sensible approach to address this. Ultimately, it is arguable that better prospective pragmatic clinical research is needed in this area to test which imaging approach is best.
In conclusion, there is evidence to support both the use of CT and radiographs in assessing scaphoid fracture union, although CT appears to be superior in terms of both its diagnostic characteristics and reliability. These findings are consistent with the trend in current practice towards a greater use of early CT, when the images obtained are of a standardized quality, taken at an appropriate time interval, and interpreted in a consistent way by those suitably trained to do so. Further research is necessary to better define the optimal clinical pathways for patients.
References
1. Leslie IJ , Dickson RA . The fractured carpal scaphoid. Natural history and factors influencing outcome . J Bone Joint Surg Br . 1981 ; 63-B ( 2 ): 225 – 230 . Crossref PubMed Google Scholar
2. Shenoy R , Pillai A , Hadidi M . Scaphoid fractures: variation in radiographic views - a survey of current practice in the West of Scotland region . Eur J Emerg Med . 2007 ; 14 ( 1 ): 2 – 5 . Crossref PubMed Google Scholar
3. Yin ZG , Zhang JB , Kan SL , Wang XG . Diagnosing suspected scaphoid fractures: a systematic review and meta-analysis . Clin Orthop Relat Res . 2010 ; 468 ( 3 ): 723 – 734 . Crossref PubMed Google Scholar
4. Dias JJ , Kheiran A , Ngo DN , et al. The onset, progress and factors influencing degenerative arthritis of the wrist following scaphoid fracture non-union . Injury . 2023 ; 54 ( 3 ): 930 – 939 . Crossref PubMed Google Scholar
5. Farracho LC , Moutinot B , Neroladaki A , et al. Determining diagnosis of scaphoid healing: comparison of cone beam CT and X-ray after six weeks of immobilization . Eur J Radiol Open . 2020 ; 7 : 100251 . Crossref PubMed Google Scholar
6. Ammori MB , Elvey M , Mahmoud SS , et al. The outcome of bone graft surgery for nonunion of fractures of the scaphoid . J Hand Surg Eur Vol . 2019 ; 44 ( 7 ): 676 – 684 . Crossref PubMed Google Scholar
7. Liberati A , Altman DG , Tetzlaff J , et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration . BMJ . 2009 ; 339 : b2700 . Crossref PubMed Google Scholar
8. No authors listed . Covidence . http://www.covidence.org ( date last accessed 11 February 2025 ). Google Scholar
9. Rossi AR , DeSmet AA , Engber WD , Tuite MJ . Plain film evaluation of bone grafting for nonunited scaphoid fractures . Skel Radiol . 1995 ; 24 ( 5 ): 361 – 364 . Crossref PubMed Google Scholar
10. Whiting PF , Rutjes AWS , Westwood ME , et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies . Ann Intern Med . 2011 ; 155 ( 8 ): 529 – 536 . Crossref PubMed Google Scholar
11. Buijze GA , Wijffels MME , Guitton TG , et al. Interobserver reliability of computed tomography to diagnose scaphoid waist fracture union . J Hand Surg Am . 2012 ; 37 ( 2 ): 250 – 254 . Crossref PubMed Google Scholar
12. Hannemann PFW , Brouwers L , van der Zee D , et al. Multiplanar reconstruction computed tomography for diagnosis of scaphoid waist fracture union: a prospective cohort analysis of accuracy and precision . Skel Radiol . 2013 ; 42 ( 10 ): 1377 – 1382 . Crossref PubMed Google Scholar
13. Hannemann PFW , Brouwers L , Dullaert K , van der Linden ES , Poeze M , Brink PRG . Determining scaphoid waist fracture union by conventional radiographic examination: an analysis of reliability and validity . Arch Orthop Trauma Surg . 2015 ; 135 ( 2 ): 291 – 296 . Crossref PubMed Google Scholar
14. Matzon JL , Lutsky KF , Tulipan JE , Beredjiklian PK . Reliability of radiographs and computed tomography in diagnosing scaphoid union after internal fixation . J Hand Surg Am . 2021 ; 46 ( 7 ): 539 – 543 . Crossref PubMed Google Scholar
15. Dias JJ , Brealey SD , Fairhurst C , et al. Surgery versus cast immobilisation for adults with a bicortical fracture of the scaphoid waist (SWIFFT): a pragmatic, multicentre, open-label, randomised superiority trial . Lancet . 2020 ; 396 ( 10248 ): 390 – 401 . Crossref PubMed Google Scholar
16. Hannemann PFW , van Wezenbeek MR , Kolkman KA , et al. CT scan-evaluated outcome of pulsed electromagnetic fields in the treatment of acute scaphoid fractures: a randomised, multicentre, double-blind, placebo-controlled trial . Bone Joint J . 2014 ; 96-B ( 8 ): 1070 – 1076 . Crossref PubMed Google Scholar
17. Dias JJ , Taylor M , Thompson J , Brenkel IJ , Gregg PJ . Radiographic signs of union of scaphoid fractures. An analysis of inter-observer agreement and reproducibility . J Bone Joint Surg Br . 1988 ; 70-B ( 2 ): 299 – 301 . Crossref PubMed Google Scholar
18. Drijkoningen T , Ten Berg PWL , Guitton TG , Ring D , Mudgal CS . Reliability of diagnosis of partial union of scaphoid waist fractures on computed tomography . J Hand Microsurg . 2018 ; 10 ( 3 ): 130 – 133 . Crossref PubMed Google Scholar
19. Geoghegan JM , Woodruff MJ , Bhatia R , et al. Undisplaced scaphoid waist fractures: is 4 weeks’ immobilisation in a below-elbow cast sufficient if a week 4 CT scan suggests fracture union? J Hand Surg Eur Vol . 2009 ; 34 ( 5 ): 631 – 637 . Crossref PubMed Google Scholar
20. Grewal R , Frakash U , Osman S , McMurtry RY . A quantitative definition of scaphoid union: determining the inter-rater reliability of two techniques . J Orthop Surg Res . 2013 ; 8 ( 1 ): 1 – 5 . Crossref PubMed Google Scholar
21. Singh HP , Forward D , Davis TRC , Dawson JS , Oni JA , Downing ND . Partial union of acute scaphoid fractures . J Hand Surg Br . 2005 ; 30 ( 5 ): 440 – 445 . Crossref PubMed Google Scholar
22. Landis JR , Koch GG . The measurement of observer agreement for categorical data . Biometrics . 1977 ; 33 ( 1 ): 159 – 174 . Crossref PubMed Google Scholar
23. Fleiss JL . Measuring nominal scale agreement among many raters . Psychol Bull . 1971 ; 76 ( 5 ): 378 – 382 . Crossref Google Scholar
24. Power M , Fell G , Wright M . Principles for high-quality, high-value testing . Evid Based Med . 2013 ; 18 ( 1 ): 5 – 10 . Crossref PubMed Google Scholar
25. Dean BJF , *On behalf of the SUSPECT study group . The management of suspected scaphoid fractures in the UK: a national cross-sectional study . Bone Jt Open . 2021 ; 2 ( 11 ): 997 – 1003 . Crossref PubMed Google Scholar
26. Brekke AC , Snoddy MC , Lee DH , Richard MJ , Desai MJ . Biomechanical strength of scaphoid partial unions . J Wrist Surg . 2018 ; 7 ( 5 ): 399 – 403 . Crossref PubMed Google Scholar
27. Rothenfluh E , Jain S , Guggenberger R , Taylor WR , Hosseini Nasab SH . The influence of partial union on the mechanical strength of scaphoid fractures: a finite element study . J Hand Surg Eur Vol . 2023 ; 48 ( 5 ): 435 – 444 . Crossref PubMed Google Scholar
28. No authors listed . Scaphoid fracture standards . The British Society for Surgery of the Hand . 2024 . https://www.bssh.ac.uk/_userfiles/pages/files/professionals/Trauma%20standards/Scaphoid%20standards.pdf ( date last accessed 11 February 2025 ). Google Scholar
29. Iordache SD , Goldberg N , Paz L , Peylan J , Hur RB , Steinmetz A . Radiation exposure from computed tomography of the upper limbs . Acta Orthop Belg . 2017 ; 83 ( 4 ): 581 – 588 . PubMed Google Scholar
30. Rua T , Gidwani S , Malhotra B , et al. Cost-effectiveness of immediate magnetic resonance imaging in the management of patients with suspected scaphoid fracture: results from a randomized clinical trial . V Health . 2020 ; 23 ( 11 ): 1444 – 1452 . Crossref PubMed Google Scholar
Author contributions
G. Smith: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Writing – original draft, Writing – review & editing
W. H. Teng: Data curation, Formal analysis, Investigation, Writing – review & editing
N. D. Riley: Conceptualization, Writing – original draft, Writing – review & editing, Investigation
C. Little: Conceptualization, Writing – original draft, Writing – review & editing, Investigation
E. Sellon: Conceptualization, Writing – original draft, Writing – review & editing, Investigation
N. Thurley: Investigation, Writing – review & editing
J. Dias: Supervision, Writing – review & editing, Investigation
B. J. F. Dean: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Validation, Writing – original draft, Writing – review & editing
Funding statement
The author(s) disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: AO UK research grant.
ICMJE COI statement
B. J. F. Dean reports an Orthopaedic Research UK grant, which is unrelated to this work. J. Dias discloses an unrelated SWIFFT NIHR grant. N. D. Riley declares consulting fees from Acumed, Arthrex, and Sovereign Medical; and being on the orthopaedic data evaluation hand and wrist panel, all of which are also unrelated.
Data sharing
The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request.
Open access funding
The open access fee for this paper was funded by an AO UK grant.
Supplementary material
Tables showing: which studies assessed which particular imaging characteristics; and the asssessment of union methods, observer details, and reference standard. A PRISMA checklist has also been included.
© 2025 Smith et al. This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives (CC BY-NC-ND 4.0) licence, which permits the copying and redistribution of the work only, and provided the original author and source are credited. See https://creativecommons.org/licenses/by-nc-nd/4.0/
