The suspected scaphoid injury: resource implications in the absence of magnetic resonance imaging

Introduction

Scaphoid fractures are common among young adults.^1,2 The average time to union for nondisplaced fractures treated in a cast is eight to 12 weeks, accounting for a considerable loss of time and productivity in this active population.^3–6 Up to one-quarter of scaphoid fractures remains undiagnosed following initial radiographs.^7–10 Undiagnosed and untreated fractures have poorer outcomes due to pseudarthrosis at the fracture site and progressive radiocarpal arthritis. As a consequence, many patients are unnecessarily immobilised for prolonged periods of time to avoid missing a small number of occult fractures.

Magnetic resonance imaging (MRI) is known to have a high sensitivity and specificity in the diagnosis of occult fractures of the scaphoid.^4,7,10–15 However, many units do not have the facilities or resources to use MRI as an early diagnostic adjunct in patients with clinically suspected scaphoid fractures.

The Royal Infirmary of Edinburgh (RIE) serves a defined adult patient population of approximately 570,000. At present, the routine use of MRI is not available for the early diagnosis of occult scaphoid injury. The primary aim of this study was therefore to investigate the epidemiology of scaphoid-related presentations to the orthopaedic fracture clinic. The secondary aims involved reporting the process of care, and calculating the costs involved, in the treatment of these common injuries.

Methods

The study was approved by the local research and ethics committee.

We prospectively identified all patients aged 13 years or older, referred from the RIE emergency department (ED) to the RIE orthopaedic fracture clinic (FC), with clinical suspicion of a scaphoid injury, between October 2007 and September 2008. Clinical notes from the ED and FC, and the accompanying radiology report, were examined retrospectively. Relevant details including patient demographics, mode of injury, time to referral, period of immobilisation, use of further imaging, number of return appointments, were recorded. Patients for whom no clinical notes could be obtained were excluded, as were patients referred to FC from other sources. Children were also excluded as they were treated at a separate paediatric institution.

Three injury groups were defined:

Statistical analysis

All data were recorded and analysed using SPSS Version 18.0 (SPSS Inc., Chicago, IL). Categorical variables were compared using chi-square tests. Independent t-tests were used to compare observed means between groups of continuous variables. A p value of 0.05 was used to determine significance.

Cost analysis

Diagnostic and treatment costs were calculated with reference to staff, materials, and general overheads. Monetary costs were derived from local hospital management data, and rounded to the nearest pound sterling (£). The cost of physiotherapy treatment and referral had to be omitted for all groups, due to a lack of reliable referral and attendance data from the physiotherapy department.

Results

After exclusions (n = 17), 537 patients with scaphoid-related presentations were referred to the FC and included in the subsequent analyses. The gender distribution was approximately equal, with 272 (50.7%) injuries seen in men. There were 87 (16.2%) TFs, 43 (8.0%) OFs, and 407 (75.8%) suspected scaphoid injuries. A significant difference in the gender-related distribution of injury groups was noted (p < 0.0001, chi-square test) (Figure 1). OPEN IN VIEWER

TFs were more commonly seen in men than in women. The mean age of men with TFs was 30.1 years (±12.9 yrs, SD), while similarly affected women were noted to be significantly older at 46.9 years (±20.9 yrs, SD) (p < 0.0001, t-test). OFs affected less than 10% of men and women, with mean patient ages of 32.4 years (±16.9 yrs) and 41.4 years (±18.9 yrs), respectively (p = NS). Suspected scaphoid injuries made up the largest injury group, and were more frequently seen in women (37.4 yrs ± 20.9 yrs) than in men (33.8 yrs ± 17.6 yrs), (p = NS).

All scaphoid-related injury occurred more often in younger adult groups (Figure 2). Patients under the age of 50 years accounted for 77.3% of patients in this study. The frequency of TFs was greatest in patients aged 20 to 29 years, while suspected scaphoid injuries were most frequent in teenagers. Retrospective review of the clinical notes revealed data on the injury mechanism for 505 (94.0%) patients. A simple fall from a standing height and sports-related accidents were responsible for the majority (79.0%) of scaphoid-related injuries (Figure 3). These findings are in keeping with previous reports. ¹⁶ OPEN IN VIEWER

OPEN IN VIEWER

Figure 3.

Mechanisms of injury (n) for scaphoid-related fracture clinic referrals.

True fractures

Eighty seven patients suffered TFs, with a gender ratio of 74:26 (M:F,%). The mean length of time between presentation to the ED and FC attendance was 2.8 days (±1.5 days). In addition to the initial set of radiographs obtained in the ED, TF patients required on average a further 1.6 (±0.7) radiographic series during their process of care pathway. Sixty seven TFs were treated definitively in cast, while 20 were treated operatively, and this decision was made by the senior orthopaedic clinician in FC. The cast group required a total of 3.4 (±1.9) FC visits. They were immobilised for a period of 7.1 weeks (±2.8 wks).

Patients in the operative group were always referred to a specialist upper limb surgeons FC, requiring a total of 4.4 (±2.3) FC visits, and remaining immobilised in cast until surgery. Fourteen patients underwent surgery primarily (by way of open or percutaneous compression screw fixation). Six patients had surgery performed at a later date (for failed cast treatment). Two patients underwent revision surgery following non-union. The average cost of treatment for TFs (treated non-operatively and operatively) was £1173 per patient (Table 1).

OPEN IN VIEWER

Table 1.

Cost analysis for true fractures and occult fractures of the scaphoid.

	Average cost per patient (£)
	True fractures
	Cast group (n = 67)	Operative group (n = 20)	Occult fractures (n = 43)
Emergency dept	164	193	176
Radiographs	20	20	20
Fracture clinic	628	778	512
Radiographs	39	45	39
Surgery	0	1210	26
Subtotal average	852	2246	–
Overall average cost	1173		773

Occult fractures

Forty three patients, with a gender ratio of 58:42 (M:F,%), were treated for an occult scaphoid fracture. The current treatment protocol in the ED suggests a delayed FC appointment, and this was reflected in the mean time from ED to FC (11.8 days ± 7.3 days). High-energy injury mechanisms were not encountered in this group. Patients injured through sporting activity were more likely to have an OF (OR = 2.46, 95% CI 1.27–4.79) when compared with the other two injury groups (p = 0.011, chi-square).

Four OF patients had the diagnosis confirmed at FC by repeat clinical examination and review of the initial ED films. Twenty four patients required further imaging (one computed tomography scan, one MRI, 22 plain radiographs) before the fracture could be identified. Of note, six patients in this subgroup had not received a full scaphoid series in the ED. In 15 patients the diagnosis was made on the grounds of continuing clinical suspicion alone, without confirmatory radiological evidence. OFs of the scaphoid therefore represented 33.1% of all fractures seen. Patients with OFs required an additional 1.6 (±0.6) sets of radiographs. The average length of immobilisation was 6.0 ± 2.4 weeks. The average cost of the treatment of OFs was £773 per patient (Table 1).

Suspected scaphoid injury

Four hundred and seven patients were found to have suffered suspected scaphoid injuries. This figure represents 90.5% of all patients whose initial radiographs were negative for fracture. Men were outnumbered by women (38:62, M:F%), and this gender distribution differed significantly from that seen in the other two injury groups (p  ≤ 0.0001; chi-square). Eighty five patients either cancelled their appointment or failed to attend FC. Of those available for review, the final FC diagnoses included soft tissue injury (n = 300), fracture of the distal radius (n = 10), fracture elsewhere in the carpus (n = 3), fracture of the first metacarpal (n = 2), rupture of the ulnar collateral ligament of the thumb (n = 2), and exacerbation of degenerative joint disease (n = 5). Low-energy injury mechanisms predominated.

Patients with suspected scaphoid injuries received a further 2.0 (±0.7) sets of radiographs in total. Twenty five (6.1%) had re-attended the ED with persistent symptoms. For those patients who attended for review, the mean time to FC attendance was 11.3 ± 6.2 days, and patients were reviewed in the FC on a further 1.6 (±0.9) occasions.

Of the 407 suspected scaphoid injury patients:

85 (20.9%) did not attend FC,

The length of immobilisation varied depending on the final diagnosis but was on average 2.9 weeks (±1.8 wks). The average cost of the treatment of suspected scaphoid injuries varied depending upon the number of fracture clinic appointments attended (Table 2).

OPEN IN VIEWER

Table 2.

Cost analysis for suspected scaphoid injuries.

No. of FC attendances	Average cost per patient (£)
	0 (n = 85)	1 (n = 195)	≥ 2 (n = 127)
Emergency dept	164	163	172
Radiographs	23	22	22
Fracture clinic	0	175	448
Radiographs	0	10	29
Surgery	0	0	0
Overall	187	371	666

Discussion

Five hundred and thirty seven patients attended the ED of the Royal Infirmary of Edinburgh with a scaphoid-related injury in one year. Only one-quarter of these patients was eventually diagnosed with a scaphoid fracture. Male gender and sporting activity were the only significant risk factors for fracture, and this has been shown previously. ¹³ The frequency of occult injuries within the fracture group was 33.1%, which is higher than previously reported.^7–10

Occult fractures and suspected scaphoid injury

Patients whose initial radiographs showed no definite fracture made up the vast majority of the study group (83.3%). Only 9.5% of these patients were eventually diagnosed with a scaphoid fracture. Indeed, 20% of this group chose not to attend their FC appointment for further review. The suspected injury group was immobilised for an average period of 2.9 weeks. In 14.4% of these patients, an alternative diagnosis was found, most commonly a fracture of the distal radius. However, this figure is likely to be an underestimate as MRI is not routinely available as an early diagnostic adjunct at our institution. In a previous prospective study (which utilised MRI scanning), the authors showed that of all injuries detected on MRI, only 25% involved fracture of the scaphoid. ¹⁷ The other 75% consisted of non-scaphoid carpal fractures (13%), avulsion fractures of extrinsic ligaments (12.5%), fractures of the distal radius (12%), bone bruises (10%), other fractures (26.5%), and soft tissue injuries (26%).

It has been demonstrated that MRI is a reliable method of diagnosing OFs of the scaphoid.^4,7,11–15 MRI has a higher sensitivity and specificity in the diagnosis of OFs of the scaphoid compared to other methods of diagnosis and might even be more cost -effective than repeated clinical examinations and plain radiographs.^4,13 Our cost analysis has estimated MRI to cost £97 per patient, yet it is seldom utilised as a diagnostic adjunct at our institution. Some authors have suggested that extremity MRI is likely to cost less. ¹⁰ One appointment at the FC costs £175 per patient. In this series, the average additional cost for patients who attend FC once (£184) or two or more times (£479) was far greater than the cost of MRI.

This study represents a large number of prospectively and consecutively recorded scaphoid-related injuries. It includes a simple comparison of the costs involved in the management of three scaphoid injury groups. Unfortunately, the retrospective nature of our case note review precluded the analysis of predictive examination findings, which has previously been of interest in the literature. ¹⁸ This study has, however, identified the cost to the patient and the cost to our department, relying solely on plain radiography and repeated clinical review in the diagnosis of occult scaphoid injuries.

The Royal Infirmary of Edinburgh does not have an imaging protocol for occult scaphoid injuries, and this is said to be the case in approximately 80% of UK hospitals. ¹⁹ Based on the results presented here, the development of an early MRI protocol in our unit is likely to improve the diagnosis of occult scaphoid and soft tissue injuries, while avoiding unnecessary repeat FC attendances, prolonged periods of immobilisation, and unnecessary expenditure.