Authors: Margaret Ting-Kwei Chang, Maggie Price, James Furness, Kevin Kemp-Smith, Vini Simas, Rowan Pickering, Deborah Lenaghan
Categories: Research Article, Observational Study
Source: Medicine
Scaphoid fractures are commonly present to emergency departments (EDs), challenging medical practitioners to achieve accurate diagnosis and management. This is because of the prevalence of radiographically occult scaphoid fractures and complications associated with missed diagnoses. Clinical Guidelines are limited for treatment of suspected scaphoid fractures, and heterogeneity in the literature further complicates management. This study aimed to explore the differences in management between practitioners in the ED and determine if immobilizing clinically suspected scaphoid fractures is supported by current evidence. This study also aimed to establish if there are predictors to assist in the diagnosis of a scaphoid fracture in the ED.
A retrospective cohort study analyzed clinical data from patient’s charts who attended the ED for a scaphoid fracture in 2019. Using retrospective patient chart audits and a Data Extraction Form, the clinical data regarding the assessment, treatment, diagnosis, and follow-up outcomes were collected. Descriptive analysis and multivariable logistic regression were performed to assess current management and find out predictors of a scaphoid fracture.
There was significance between practitioners performing physical assessments and providing treatment (P < .001). Physiotherapists performed assessment and education combined treatment more frequently than nurse practitioners and doctors. Thirty-four cases (11.7%) were negative for fracture in ED and positive in follow-up at the orthopedic clinic. There was an estimated loss of income of $327,433.60 (Australian dollar) for 221 patients who missed work due to overtreatment with immobilization. The strongest predictors for a confirmed scaphoid fracture were of male gender (odds ratio, 3.2; 95% confidence interval, 2.1–5.0; P < .001) and a positive x-ray in ED (odds ratio, 36.6; 95% confidence interval, 17.4–77.0; P < .001).
Management of scaphoid fractures across the Gold Coast Hospital Health Service ED followed commonly accepted practices involving x-ray and immobilization; however, this conservative approach to management is associated with increased health costs and low rates of conversion to a confirmed scaphoid fracture. Male gender was the only significant predictor associated with a scaphoid fracture.
Scaphoid fractures are the most common injury of the carpal bones with the literature reporting varying rates of between 50% and 80% of all carpal fractures.^[1,2]^ They are the second most common fractures to occur in the upper extremity following the distal radius. Typically, the mechanism of a scaphoid fracture is falling on to an outstretched hand (FOOSH) with an extended wrist.^[1]^ Considering the anatomy of the scaphoid bone (in its shape, orientation, and retrograde blood supply), sustaining a fracture that is not appropriately identified can lead to complications including nonunion and avascular necrosis.^[1,3]^ Eighty percent of fractures are at the waist of the scaphoid, which separates the proximal pole from the blood supply.^[1,4,5]^ Given the complex anatomy of the scaphoid bone and the associated complications of sustaining a fracture, there appears to be inconsistency among the diagnosis and management when reviewing clinical guidelines.^[6]^
While the literature presents inconsistency when diagnosing and managing radial-sided wrist pain, including suspected scaphoid fractures, variability in patient outcomes is evident and can lead to detrimental downstream effects including difficulty with return to work or everyday activities.^[3,6]^ Of all suspected scaphoid fractures, only 5% to 10% are confirmed fractures.^[7]^ Undertreatment of scaphoid fractures can lead to increased complications, such as nonunion, and overtreatment is related to unnecessary cost, such as lost workdays and medical treatment expenses.^[3]^ Previous literature has reported differences in management between practitioners even within the same hospital.^[2,3,6,8]^ It could be postulated that the inconsistency in treatment by hospitals and practitioners resulted from the lack of existing guidelines to specify appropriate treatment in the emergency department (ED).
Given the heterogeneity of the literature, it may provide some rationale as to why there is a lack of consistency in treatment between and within hospitals. Due to the risk of poor outcomes for missed fractures, which can result in chronic wrist pain and long-term disability,^[3,6]^ a highly conservative approach is used in the Gold Coast Hospital Health Service (GCHHS, Gold Coast, QLD, Australia) involving immobilization using a cast or splint for 10 to 14 days while awaiting re-x-ray.
The significance of this article is to determine if scaphoid fractures are appropriately managed in the GCHHS ED based on current evidence available, leading us to develop four primary aims of this current study. First, outline the current management approaches for suspected scaphoid fractures across all practitioners (nurse practitioners [NPs], physiotherapists [PTs], and doctors) from the GCHHS ED; benchmark the current management approaches in the ED for suspected scaphoid fractures against the highest level of evidence available; and include the conversion rates of suspected scaphoid fractures to confirmed fractures on imaging 10 to 14 days later in the Orthopedic Fracture Clinic (OFC). Second, use quality indicators from Health Analytics (a data collection program used in GCHHS) to outline patient management flow of suspected scaphoid fractures using the following National Emergency Access Target (NEAT), length of stay (LOS), time to be seen (TTS), and representation rate (i.e., patients who re-present themselves to ED for any reason). Third, determine the estimated economic impact with conservative management for suspected scaphoid fractures in the ED and OFC. Finally, determine if there are common predictors in ED that can improve detection of suspected scaphoid fractures that were confirmed scaphoid fractures in the OFC.
The researchers of this study hypothesized that the scaphoid cluster test would be highly sensitive to identify a confirmed fracture in the ED and a diagnosis in ED would be based on x-ray results regardless of possibly having an occult fracture in the acute phase. A secondary hypothesis was that management of a suspected scaphoid fracture would vary between practitioners and was generally treated conservatively in the ED due to the lack of current clinical guidelines.
This retrospective cohort study included patients who presented to the ED in 2019 with a scaphoid fracture at the Gold Coast University Hospital or Robina Hospital, two hospitals within GCHHS in Gold Coast, Australia.
Inclusion criteria were any patient diagnosed and coded in their medical record with a scaphoid fracture and patients who were treated with conservative intervention in the ED. There was no limit on age or gender of patients in this study. Exclusion criteria were any coded cases that were not recorded into electronic medical records (EMR), or integrated electronic medical records (iEMR); patients re-presenting to the ED for the same scaphoid injury; cases that were coded incorrectly and were not a scaphoid fracture diagnosis; patients who lived out of the Gold Coast area and were referred to a different hospital for follow-up; patients who were injured outside of the Gold Coast area and were initially treated in a different hospital ED; and patients who first presented to the ED for a scaphoid fracture prior to 2019. Figure 1 is a flow diagram demonstrating the process of patient selection and the number of cases included and excluded from this study.
Figure 1. Flow diagram of the participant selection included in the study. ED = emergency department, EMR = electronic medical record, iEMR = integrated electronic medical record.
The study sample size was calculated prior to initiating the study. This was calculated using a consecutive sampling approach using the Australian Bureau of Statistics sample size calculator.^[9]^ It was based on the confidence level of 95%, the population of the Gold Coast,^[10]^ the proportion of 0.5 (50%) for a conservative estimate of variance, and the confidence interval of 0.05 (5%). Based on the above calculation, the minimum sample size was a total of 384 patient charts. If this could not be achieved in the year 2019, the study inclusion would increase to 2018.
The data were obtained through retrospective patient chart audits from EMR, iEMR, and Health Analytics through GCHHS and Queensland Health. A total of 493 patient charts coded for scaphoid fracture in 2019 were accessed and screened for inclusion and exclusion criteria. Data collection was administered by two PT students (M.T.-K.C., M.P.) under the supervision of senior researchers (J.F. & K.K.S.), and GCHHS PTs (D.L. [Extended Scope] and R.P.). Included charts were reviewed over a 3-week period in sequential order based on the date of presentation in the ED. Reliability was achieved when data were cross checked between the two PT students, and any disagreements were discussed and resolved with the senior researchers and PTs at GCHHS. This study was in accordance with the Reporting of studies Conducted using Observational Routinely—collected Health Data statement under the Strengthening The Reporting of Observational Studies in Epidemiology statement.^[11,12]^ The project was made up of a single component where retrospective data were extracted and analyzed.
Data extraction followed a coded system, as shown in the Supplemental Digital Content (Appendix 1, http://links.lww.com/MD/G886). The Data Extraction Form was developed in collaboration with the GCHHS PTs to fulfill the aims of this study based on the data available from Health Analytics, EMR, and iEMR. A pilot data extraction form was used prior to starting the extraction process to minimize selection bias. All extracted data were deidentified by using an alternative numerical identifier and stored into a separate Microsoft Excel (Office 365, Microsoft Corporation, Redmond, WA) spreadsheet.
The patient’s age, gender, and hospital hospital, treating practitioner (NP, PT, or doctor), LOS, TTS, and NEAT score were collected from Health Analytics. The NEAT score was recorded as “yes” or “no” to quantify the proportion of patients who are managed and discharged within four hours of presentation to ED.^[13]^ The mechanism of injury (MOI), physical assessment, radiological assessment, diagnosis, treatment in ED, and follow-up treatment or further assessment were collected from clinical documentation in EMR and iEMR. A numerical coding system was used to categorize the data from EMR and iEMR for further statistical analysis.
The study protocol was assessed by the chair of the GCHHS Human Research Ethics Committee and approved as a Quality Activity and Clinical Audit (LNR/2019/QGC/57334). The study was not determined to be research, rather a quality activity as data was already collected as part of routine management of scaphoid injury. The consent process was not applicable as the data were collected according to routine procedures.
Data were transferred to the Statistical Package for the Social Sciences (Software version 25.0 IBM, SPSS Inc., New York, NY) for analysis where statistics such as frequencies, means, standard deviations, and ranges were presented.^[14]^ A test for normality was undertaken to determine if any skewed or normal trends for continuous variables in the 2019 year existed. Descriptive statistics were analyzed using mean and standard deviation for normally distributed variables. A descriptive analysis was conducted to present current management of suspected scaphoid fractures. The chi-square test was used for categorical variables and expressed as observation counts (and percentages). A statistician was consulted on logistic regression analysis. Simple and multivariable logistic regression was performed to determine common predictors of confirmed scaphoid fractures. Statistical significance was accepted when P values ≤ 0.05.
For cases that were lost to follow-up, relevant available data were included for our analysis, but excluded for calculations that would affect observations over a span of time. For this study, patients lost to follow-up were excluded in the chi-square test for x-ray results after 2 weeks, and the regression equations for analyzing predictors.
Table 1 outlines patient demographics of those who presented to the ED with a suspected scaphoid injury. Due to the number of patients coded for a suspected scaphoid, fracture between Gold Coast University Hospital and Robina Hospital is n = 224 and n = 225, respectively, the values in the table represent patients from both hospitals combined. The ages of the patients ranged from 8 to 90 years, although the interquartile range was between ages 14 to 41.
Figure 2 illustrates the difference between the outcome of the x-ray imaging results when it was used to confirm a diagnosis of a scaphoid fracture in the ED and the OFC. There is an estimated seven to 14 days between initial presentation in the ED and follow-up in the OFC where the patient’s wrist was immobilized. During this time period, 64 patients were recorded to have failed to attend their follow-up. In 290 cases that did not have a fracture in the ED, there were 34 conversions to a confirmed fracture in the OFC. That is, the conversion rate of patients presenting with a negative x-ray in the ED but having a positive x-ray in OFC is 11.7% (34/290). The x-ray results diagnosed a fracture correctly (either positive or negative) in the ED 82.1% of the time out of the 385 cases that attended a follow-up in OFC. Other fractures represented fractures that were initially coded as a scaphoid fracture, but resulted as a different diagnosis, such as distal radius or other carpal bone fracture with further investigation.
Figure 2. Flow diagram of patient x-ray results from the ED to the Orthopaedic Fracture Clinic. ED = emergency department.
Figure 3 presents the different physical assessment tests performed in the ED for determining a scaphoid fracture. Anatomical snuffbox (ASB) tenderness was the most frequently used test in the ED on the 449 patients. Common tests such as scaphoid tubercle tenderness (STT) and axial compression were fifth and sixth most frequently used among all the practitioners.
Figure 3. Physical assessments performed in the emergency department, n (%). AROM = active range of motion, ASB = anatomical snuffbox, DR = distal radius, DU = distal ulna, RD = radial deviation, UD = ulnar deviation.
Table 2 demonstrates the differences between practitioners in use of scaphoid cluster tests, treatment, and representation rates. The chi-square test showed a significant difference between practitioners using the scaphoid cluster with patients (x^2^ = 22.1, P < 0.01). The treatment cluster represented the combined use of the top three treatments (cast/backslab, simple analgesia, and education), and a chi-square test reported significant difference between practitioners (x^2^ = 151.2, P < 0.01). These treatments were chosen for the cluster as they demonstrate the basic treatment for conservative management with immobilization, pain relief, and appropriate advice and education. There was no significant difference in the rate of representation to the ED between practitioners (x^2^ = 1.0, P = 0.6).
Figure 4 demonstrates the different treatments used in the ED for scaphoid fractures. Immobilization with a cast or backslab was used in 97.1% of all patients. Most patients received a combination of cast, simple analgesia (such as ibuprofen, paracetamol) and education including rest, ice, compression, and elevation (RICE) advice. There was one case of no treatment documented.
Figure 4. Treatment data collected from EMR and iEMR. EMR = electronic medical records, iEMR = integrated electronic medical records.
Figure 5 presents the number of patients working or not working at the time of their injury. Sixty-eight patients did not have a work status recorded; therefore, it was assumed that patients between the ages of 18 and 65 were working within this group. A total of 221 patients were considered to be working and had lost income due to being immobilized.
Figure 5. Number of patients who were working or not working at the time of their injury.
Table 3 presents the estimated loss of income for the 221 patients over the 10 to 14 days of immobilization while waiting for a confirmed scaphoid fracture diagnosis on re-x-ray at their follow-up in the OFC. The estimated average cost of OFC appointments based on GCHHS data over a 12-month period from 2017 to 2018 is also presented. The minimum wage in Australia was 740.80 per week as reported by Fair Work Ombudsman and was used for a conservative estimate to calculate the economic impact of wages lost.^[15]^ The economic impact from health care costs was also considered in relation to the number of appointments needed to manage a suspected scaphoid fracture. The average cost of OFC appointments per patient was $893.00 for 4 to 5 total visits.
Positive tests for ASB tenderness, STT, and axial compression were significant when analyzed individually; however, when tested as a cluster, it was not significant (Table 4). This study found the odds of having a confirmed scaphoid fracture on the follow-up x-ray was reduced by 0.5 with a positive ASB test in ED, or 49.4% less likely of having a fracture with a positive test. The same was true for STT, axial compression, and cluster tests where the odds ratio (OR) was reduced for a fracture confirmed at follow-up with a positive ED test.
There was a significant difference in the proportion of males versus females who were diagnosed with a confirmed scaphoid fracture outcome (Table 5). The odds of having a confirmed scaphoid fracture was 3.2 times higher in males than females. Age groups 15 to 40 and under 14 were not significantly different when compared to those aged 40 to 90, although there was a large portion of patients who presented to the ED that were under the age of 40. Injury of the dominant hand compared to the nondominant hand was not a significant predictor for a scaphoid fracture.
The odds of having a confirmed scaphoid fracture in the OFC was significantly greater with a positive x-ray in the ED at 36.6 times higher (Table 6). A negative x-ray in ED had a significantly reduced risk of a fracture in the OFC by 97.3%. The MOI did not have a significant difference in the fracture outcome; however, FOOSH was 1.7 times higher than a different MOI.
The best model for predicting risk of a confirmed scaphoid fracture was the presence of a positive x-ray and being the male gender. The odds of a confirmed scaphoid fracture were 33.7 times higher with a positive x-ray, and 2.7 times higher in males when FOOSH and 1 positive cluster test was also present (Table 7). Although FOOSH and 1 positive cluster test were not significant, the OR increased the risk of having a confirmed scaphoid fracture by 83% and 19%, respectively.
The aims of this study were to benchmark GCHHS practice with current evidence and differences between practitioners, appraise management flow using quality indicators, evaluate the economic impact of current management, and determine if there were any predictors to improve the diagnosis of scaphoid fractures in the ED. Across the GCHHS, the authors found that there were differences between Practitioners (NP, PT, and doctors) in the delivery of assessment and treatment for a suspected scaphoid fracture.
To compare the differences between practitioners performing a physical assessment, a scaphoid cluster test (ASB tenderness, STT, and Axial Compression) was formulated based on the recommendations of Mallee et al ^[16]^ and Carpenter et al.^[17]^ The scaphoid cluster test in this study showed some positive intent (OR, 0.6; 95% confidence interval, 0.3–1.1); however, it was not statistically significant. A systematic review and meta-analysis compared the specificity and sensitivity of several scaphoid tests, reporting a significant heterogeneity in the test results when used individually.^[16]^ Mallee et al ^[16]^ found that combining tests improved the sensitivity (1.00) and specificity (0.74), and a 64% probability of having a scaphoid fracture if all three tests were positive ensuring that a fracture will not be missed.^[16]^ Carpenter et al ^[17]^ observed similar results but used the tests to rule out a fracture. Given the high sensitivity and specificity of the scaphoid cluster test, it would seem appropriate to analyze the accuracy of assessment in this study and encourage this type of testing in practice. Fewer reports of the cluster test being documented by some practitioners (18.3%, 82/449) may have influenced the strength of our results and, therefore, should be interpreted with caution.
Overall, the result indicated there were several assessments performed by practitioners when managing scaphoid fractures. In the current study, the results showed a significant difference between practitioners using the cluster test when assessing patients (x^2^ = 22.1, P < .001). Although doctors assessed the greatest number of patients (n = 195), PTs utilized the cluster tests in their assessment the most often. Even so, PTs only used all three tests in this cluster in 28.6% of the 175 cases. Previous literature reports differences in management between practitioners within the same hospital, which led to a lack of consistency in assessment outcomes.^[2,3,6,8]^ This variability in approaches across practitioners in the current study may be a reflection of the lack of clear recommendations in the available clinical guidelines. These guidelines do not include recommendations for specific assessment protocols of a clinically suspected scaphoid fracture in the ED.^[2,6]^
Practitioners appear to be more systematic in the delivery of treatment than in assessment, but it is evident that there is variability between practitioners in management protocols for a suspected scaphoid fracture. Most practitioners treated patients following standard practice for fracture management, which included a combination of immobilization, prescribed simple analgesia, and provided education (treatment cluster). Results of the current study indicated that there was a significant difference between practitioners using the treatment cluster. PTs used this cluster most often in 82.3% of the patients they treated.
It is widely accepted in literature that conservative treatment includes immobilization of the scaphoid until the fracture can be radiologically confirmed 10 to 14 days after injury.^[8]^ While this notion of immobilization is widely accepted, there is a lack of evidence supporting the amount of immobilization. Previous research has shown that various forms of casting (above or below elbow) for conservative treatment revealed no significant difference in rates of nonunion between different types of casts, indicating that greater immobilization had no impact on healing.^[18]^ As previously mentioned, the lack of clinical guidelines available may contribute to the variability in treatment approaches between practitioners. It can be postulated that the lack of a consistent protocol in assessment impacts the decisions to determine the best treatment approach, which can result with inconsistencies such as overtreatment or unnecessary costs.
In this study, a conversion rate was calculated to determine the number of clinically suspected scaphoid fractures in the ED with a negative x-ray image, which resulted in a radiologically confirmed scaphoid fracture in the OFC. Within the 290 cases that presented with a negative x-ray in ED, 34 (11.7%) cases were confirmed at follow-up. This percentage is in keeping with previously reported literature^[7]^ reporting a 5% to 10% conversation rate of negative to positive imaging in the follow-up appointment. The literature shows the probability of a scaphoid fracture is 25%, based on history, and physical examination following initial x-rays without an apparent fracture.^[17]^ These reported figures have driven the shift to early immobilization in cases of a suspected scaphoid injury in the presence of a negative x-ray. Conversely, previous studies have found about 75% of patients with a clinically suspected scaphoid fracture did not have a fracture and were unnecessarily immobilized.^[19,20]^ There is a possibility that this number can be reduced if patients were referred for magnetic resonance imaging (MRI) rather than an x-ray during their initial assessment. The literature recommends using MRI for an early and accurate diagnosis of a scaphoid fracture over computed tomography, x-ray, and bone scintigraphy.^[3,17,21–23]^ The literature also consistently shows MRI is superior in sensitivity and specificity over all other advanced imaging devices.^[3,17,21–23]^
Across hospitals in Australia, NEAT, LOS, and TTS are key indicators for quality of care and timeliness within the ED. Wait times are a major concern in the overall quality of care provided in health care.^[13,24]^ In this study, GCHHS met the NEAT criteria 87% of the time with a median LOS being 119 minutes and 14 minutes for TTS. The Australian Institute of Health and Welfare reported that 71% of all ED presentations across Australia were completed within the four hour stay NEAT criteria in 2017 to 2018 report.^[25]^ The utilization of NP and PT practitioners for minor musculoskeletal injuries in the ED likely contributes to the high level of care provided in GCHHS when managing scaphoid fractures as they met these targets within both hospitals.^[26,27]^ This study demonstrated a representation rate of 13% with the majority of cases representing due to problems with their cast (e.g. pain, discomfort, ill-fit, or damaged). Representation rate is not included in the quality indicators across Australian EDs, and there is currently no strong evidence that representation impacts outcomes.^[24]^ Therefore, continued efforts for reducing LOS and TTS would increase overall quality of care.^[13]^
The current standard practice follows a conservative approach, which tends to provide overtreatment and unnecessary costs.^[3,8]^ This study demonstrated that 239 patients received treatment favoring conservative approaches against a possible missed diagnosis and had increased direct health care costs including immobilization materials, repeated imaging, and repeated hospital attendance. This conservative approach also resulted in indirect costs including missed workdays and alternative transportation or missed work days of family members to transport the patient to attend OFC appointments since patients are unable to drive while they are in a cast. Of the 449 patients who presented to the ED, 221 were of working age or reported to be working at the time of their injury. An estimated minimum total of 213,427.00 (AUD). This is an estimated total economic impact of $540,860.60 (AUD). There have been several studies that have completed cost-effective analysis on standard practice versus early MRI for diagnosis, finding that the use of MRI is beneficial as it decreases both direct and indirect health costs for suspected scaphoid fractures.^[8,20]^ The National Institute for Health and Care Excellence Guidelines suggest that MRI for diagnosis is a more cost-effective investigation that allows more accurate diagnosis earlier and reduces unnecessary costs associated with overtreatment.^[22]^
To determine if there were common predictors that can improve early detection of a scaphoid fracture in ED, data analyses using simple and multivariate logistic regression equations were used. This tested the probability of various predictors that would suggest a scaphoid fracture was likely on follow-up x-ray in the OFC. In a multivariate regression equation, this study showed that the strongest combination of characteristics to predict a confirmed scaphoid fracture included male gender, FOOSH, 1 of 3 positive scaphoid cluster tests, and a positive x-ray in the ED. With this combination of predictors, only male gender and a positive x-ray in the ED were statistically significant. Although FOOSH and having 1 of 3 positive scaphoid cluster tests were not significant, it strengthened other predictors in the equation. This is supported by literature in regard to male gender as a predictor.^[16,28]^ Although some articles suggest that a younger population is also a predictor, it was not statistically significant in our analysis. The scaphoid cluster test independently is not a predictor of scaphoid fracture; however, it may still be clinically useful when combined with x-ray in ED. All scaphoid cluster tests (ASB tenderness, axial compression, and STT) have high sensitivities (0.96, 0.82, and 0.92, respectively) and low specificities (0.39, 0.58, and 0.47, respectively) making it a good predictor to rule out a scaphoid fracture if all tests are negative.^[16,17]^ Additionally, x-ray results have a higher specificity (0.72) and relatively lower sensitivity (0.82) than the physical tests, suggesting it is a good predictor to rule in a scaphoid fracture diagnosis.^[17]^
Further research is recommended to refine the standard management practices for suspected scaphoid fractures, to determine how the male gender is an influencing predictor that can play a role in the management of clinically suspected scaphoid fractures, and to determine when immobilization in the presence of a negative x-ray is appropriate. Future research from this study could include evaluating the use of a brace over a cast for clinically suspected scaphoid fractures and compare the rates of nonunion or complications between these two interventions.
A standardized protocol could benefit practitioners to prevent overtreatment and undertreatment of scaphoid fractures. Some practices suggested by the authors of this study could
This study was performed retrospectively and is level III-2 evidence and should be interpreted cautiously.^[30]^ The sample from this study only included two public hospitals on the Gold Coast. In order to extend these findings across EDs in Australia, similar studies should be designed.
In regard to the patients who have been lost to follow-up after presenting to ED, it was not documented in their medical records whether they had received further imaging or management privately for the scaphoid fracture versus no further treatment.
In data extraction, the authors did not record the number of days between discharge from ED to the first appointment of the follow-up in OFC, nor was it recorded the number of times the patient attended follow-up appointments after the first follow-up attendance. This data could have further informed the economic impact.
For the patients with a confirmed diagnosis, it was not documented to determine if their follow-up appointments continued until full recovery. Therefore, it is difficult to determine the long-term effects of treatment and if there are further complications, such as chronic disability.
For documentation of the physical assessments, it is likely that not all physical examinations performed were documented. It is possible that chart notes documented only positive physical tests that the clinician found to be a significant. This makes it difficult to collect data on negative assessment results if it was not documented. Additionally, with regard to being a retrospective study, there was no training on how the objective tests were conducted between practitioners. This may reflect potential variability in test results.
This study only had access to data of patients who presented to the ED and were coded in iEMR and EMR as a scaphoid fracture. It did not include those who presented to the ED and may have been misdiagnosed or had a missed scaphoid fracture.
In conclusion, this study found that there was a significant difference between practitioners in performing the scaphoid cluster and treatment cluster. However, this did not influence fracture outcomes as all assessments were combined with x-ray imaging in ED. The number of converted scaphoid fractures from radiographic negative to positive (11.7%) in this study was consistent with previous literature. Loss of income is an indirect health care cost that must be considered when managing suspected scaphoid fractures. Overtreatment is linked with costs in the hospital as well as for the patient. Improved practices in diagnosis would assist in decreasing direct and indirect health care costs. The outcome of initial x-ray in ED was found to be the best predictor for a scaphoid fracture. Male gender was the second-best predictor for confirmed scaphoid fractures. It was found that GCHHS followed the accepted standard practices in diagnosis and management of scaphoid fractures in ED, and that GCHHS met targets set for all Australian hospitals in NEAT, LOS, and TTS. Current guidelines recommend use of MRI for improved diagnosis, however this is not the standard practice across hospitals worldwide. Further studies are needed in acute scaphoid fractures to improve early identification and reduce unnecessary immobilization.
There were no other members outside of the included authors who have contributed to this study.
Maggie Price, Email: maggie.price@student.bond.edu.au.
James Furness, Email: jfurness@bond.edu.au.
Kevin Kemp-Smith, Email: kkempsmi@bond.edu.au.
Rowan Pickering, Email: rowan.pickering@health.qld.gov.au.
Deborah Lenaghan, Email: deborah.lenaghan@health.qld.gov.au.