Epidemiology and Clinical Implications of Scaphoid Morphology Based on Morsy Classification Authors: Chloe C. Krasnoff, Brynn A. Hathaway, Angelo B. Lipira Published: 2026 ABSTRACT PurposeA recent anatomical study in 2019 introduced a classification system of scaphoid morphology, characterizing dimensions, vascular patterns, and locations of nutrient vessels. This article aims to determine the prevalence of these scaphoid types in the general population.MethodsA retrospective, single-center study was performed. Patients who underwent wrist imaging via 4-view wrist x-rays from January to December 2022 were included. Exclusion criteria included patients with acute scaphoid fracture or pathology. Anatomic scaphoid measurements were performed for each radiograph. Chi-squared tests and unpaired t tests were used for categorical variables and quantitative variables, respectively. Statistical significance was noted for P values <.05.ResultsThree hundred sixty-three x-rays from 345 patients met the inclusion criteria, including 291 type I patients (80.2%) and 72 type II patients (19.8%). Patients with type II scaphoids were significantly more likely to be left handed (P < .05); otherwise, there were no demographic differences between type I and II patients.ConclusionsType I scaphoids are more prevalent in the general population, 80% in this study. No clinically significant demographic or patient factor differences were found between patients with type I and II scaphoids. Future studies are warranted to determine the clinical impact of type II scaphoids and the development of scaphoid-related pathology (avascular necrosis, fractures, nonunions, etc).Type of study/level of evidence2c ecologic study The scaphoid is the most commonly fractured carpal bone, accounting for 50% to 70% of carpal fractures.1,2 The scaphoid’s unique anatomy, variable shape, and blood supply have been extensively studied.2, 3, 4, 5 These anatomic considerations impact multiple clinical entities, including scaphoid nonunion, avascular necrosis (including Preiser disease), and scaphoid nonunion advanced collapse.6 An understanding of morphology and blood supply may inform surgical management of scaphoid fractures and impact prognosis. In a 2019 cadaver study (N = 13) using microcomputed tomography, Morsy et al1 proposed two distinct scaphoid morphologies (types I and II) based on specific length and width ratios for scaphoids on x-ray. This included the length of the capitate fossa in a posteroanterior view with ulnar deviation, and the narrowest width at the waist in the navicular view. These expressed as ratios to the longest length of the scaphoid were termed the capitate fossa index (CFI) and waist index (WI) (Fig.).1 This study demonstrated differences in both the number and location of nutrient vessels between the two groups. Type I scaphoids were found to have a more robust vascular network compared with type II scaphoids.1 This finding suggests clinical implications, but to date, the prevalence of these morphological types in the general population has not been studied, nor have clinical associations been investigated.FigurePhoto from Morsy et al1 demonstrating measurements on navicular view A and lateral view B of wrist x-rays for type I and II scaphoids, including capitate fossa index (CFI) and waist index (WI). The yellow line signifies CFI, and the blue line signifies WI. Photo from Morsy et al1 demonstrating measurements on navicular view A and lateral view B of wrist x-rays for type I and II scaphoids, including capitate fossa index (CFI) and waist index (WI). The yellow line signifies CFI, and the blue line signifies WI. The primary aim of this study was to determine the prevalence of type I and II scaphoid morphologies in the general population via x-ray measurements. Secondarily, clinical and demographic associations were investigated. Methods After obtaining institutional review board approval, a retrospective single-center study was performed including consecutive patients from January 1, 2022, to December 31, 2022, who underwent 4-view x-rays of the wrist for any indication at our institution’s emergency department. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of OHSU (no. 00025272) on January 3, 2023, with the need for written informed consent waived. For inclusion, adequate posteroanterior, oblique, lateral, and navicular views of the wrist must be available. Specifically, an adequate navicular view was defined as sufficient ulnar deviation present to prevent superimposition and allow an unobstructed view of scaphoid articular surfaces, whereas an adequate lateral view was defined as a relationship between the palmar cortex of the pisiform being within the scaphoid and capitate and collinear relationship between radius, lunate, and capitate. Patient information was collected including gender, age, race, handedness, diabetes, osteoarthritis of the hand/wrist, rheumatoid arthritis, osteoporosis, tobacco use, obesity (BMI > 30 kg/m2), prior surgery to the involved extremity, and presence of hand/wrist fracture to the involved extremity. Exclusion criteria included any patient with acute scaphoid pathology (eg, fracture and scapholunate injury) due to altered anatomy secondary to trauma that would affect radiographic measurements. Each patient’s x-rays were reviewed by two authors (CCK and BAH) to determine measurements and classifications based on the Morsy classification, which defined type I or II scaphoids based on two key measurements. The CFI was defined as the ratio of the length of the capitate fossa to the longest scaphoid length in a navicular view, while the WI was defined as the ratio of the narrowest width at the scaphoid waist to the longest scaphoid length in the lateral view (Fig.). They defined type I (or “full”) scaphoids having CFI < 0.6 and WI > 0.4 and type II scaphoids having CFI > 0.6 and WI < 0.4. Univariate analysis was performed using unpaired t tests for continuous variables and with Chi-squared tests for categorical variables. Fisher exact tests were used for contingency tables with cells containing values <5. Interrater reliability was determined using Cohen kappa coefficient between raters with respect to the classification of scaphoid types (type I vs type II). The following guidelines were used to describe agreement with respect to interrater reliability: slight (0.01–0.2), fair (0.21–0.4), moderate (0.41–0.6), substantial (0.61–0.81), and near perfect (>0.81). The primary outcome for agreement was the binary distinction between type I and type II, rather than agreement on the exact measurements. Results Prevalence of scaphoid types Interrater reliability for scaphoid classification (type I vs type II) demonstrated a near-perfect level of agreement (Cohen kappa, 0.86; 95% confidence interval, 0.76–0.96) on review of 50 x-rays. In total, 363 wrist x-rays from 345 unique patients (18 bilateral wrists imaged) met the inclusion criteria. The Table demonstrates demographics and measurements of the subjects. Overall prevalence included 291 wrists with type I scaphoids (80.2%) and 72 wrists with type II scaphoids (19.8%). Of the 18 patients with bilateral wrist imaging, 13 (72.2%) had the same scaphoid type on both wrists, whereas 5 (27.8%) had different scaphoid types on each wrist. Seventy-four wrists (20.4%) had concomitant fractures not involving the scaphoid, including 47 distal radius, 12 metacarpal, 12 distal ulna, 8 triquetrum, 3 proximal phalanx fractures, 1 trapezium, 1 lunate, 1 proximal ulna, and 1 proximal radius. The median values for CFI and WI were found to be 0.52 (0.08) and 0.46 (0.07) for type I scaphoids versus 0.62 (0.08) and 0.39 (0.08) for type II scaphoids.TablePatient demographics of imaged wristsPatient DemographicsType 1 (N = 291)Type 2 (N = 72)P ValueCapitate fossa index (CFI)0.52 (0.08)0.62 (0.08)Waist index (WI)0.46 (0.07)0.39 (0.08)Male gender122 (41.9)33 (45.8).36Age46 (30–61)47 (32–63.5).44Caucasian race212 (72.9)49 (68.1).66Handedness.0001 Right-handed190 (65.3)47 (65.2) Left-handed20 (6.9)2 (2.8) Ambidextrous0 (0)1 (1.3)Wrist imaged.11 Dominant hand114 (54.5)21 (42) Nondominant hand95 (45.5)29 (58)Diabetes86 (29.6)18 (25.0).92Osteoarthritis103 (35.4)30 (41.7).32History tobacco use144 (49.5)32 (44.4).44Rheumatoid arthritis10 (3.4)0 (0).11Osteoporosis23 (7.9)6 (8.3).90BMI > 30 (kg/m2)151 (51.9)42 (58.3).33Prior ipsilateral upper-extremity surgery31 (10.7)7 (9.7).82Other hand fracture65 (22.3)9 (12.5).06Bold values are statistically significant P < .05. TablePatient demographics of imaged wristsPatient DemographicsType 1 (N = 291)Type 2 (N = 72)P ValueCapitate fossa index (CFI)0.52 (0.08)0.62 (0.08)Waist index (WI)0.46 (0.07)0.39 (0.08)Male gender122 (41.9)33 (45.8).36Age46 (30–61)47 (32–63.5).44Caucasian race212 (72.9)49 (68.1).66Handedness.0001 Right-handed190 (65.3)47 (65.2) Left-handed20 (6.9)2 (2.8) Ambidextrous0 (0)1 (1.3)Wrist imaged.11 Dominant hand114 (54.5)21 (42) Nondominant hand95 (45.5)29 (58)Diabetes86 (29.6)18 (25.0).92Osteoarthritis103 (35.4)30 (41.7).32History tobacco use144 (49.5)32 (44.4).44Rheumatoid arthritis10 (3.4)0 (0).11Osteoporosis23 (7.9)6 (8.3).90BMI > 30 (kg/m2)151 (51.9)42 (58.3).33Prior ipsilateral upper-extremity surgery31 (10.7)7 (9.7).82Other hand fracture65 (22.3)9 (12.5).06Bold values are statistically significant P < .05. Patient demographics of imaged wrists Bold values are statistically significant P < .05. Analysis of demographic differences Our results showed significantly more left-hand dominant patients had a type II scaphoid (P < .0001). There were no significant differences in gender, age, race, diabetes, osteoarthritis, rheumatoid arthritis, osteoporosis, tobacco use, obesity (BMI > 30), prior surgery to the involved extremity, extremity imaged (dominant or nondominant hand), or presence of another hand/wrist fracture to the involved extremity (P > .05). Discussion This study applied the Morsy classification of scaphoid morphology to x-rays of 387 patients to estimate the prevalence of scaphoid type in the general population. Our results support that there are indeed two distinct morphological groups (type I and II scaphoids) in the general population. Similar to Morsy’s findings of 84.6% type I scaphoids versus 15.4% type II scaphoids in their 13 cadaver specimens,1 our data showed an overall prevalence of 80.2% type I scaphoids versus 19.8% type II scaphoids in a much larger sample (N = 345 patients and 363 extremities). This epidemiology is helpful for hand surgeons to be aware of, as type II scaphoids have a more slender shape and less robust vascular supply.1 The clinical impact, however, remains to be determined. Minimal differences were found between patients with type I versus type II scaphoids. This included demographic factors (eg, gender, age, race), comorbidities (eg, diabetes, osteoarthritis, tobacco use), and potential risk factors (eg, prior hand surgery, prior hand fractures). The only statistically significant finding was a higher rate of left-hand dominance in patients with type II scaphoids (P < .05). The clinical impact of these differences is likely minimal; therefore, based on our data, there are no clinically relevant demographic differences that can be used to distinguish type I versus type II scaphoids. Morsy et al’s original article outlined various nuances regarding scaphoid vascularity. In analyzing the vessel anatomical entry site, course within the scaphoid, diameter, proximal versus distal distributions, volume of scaphoid covered by distribution, and presence of anastomosis, they were able to find a clear distinction between type I versus type II scaphoids. Type II slender scaphoids had significantly fewer average number of intraosseous vascular branches, as well as fewer vessels entering dorsally, no volar waist vessels, no anastomotic network between proximal and distal blood supplies, and no vessel entering at the scapholunate ligament attachment.1 Their theoretical concerns therefore were potential for increased risk of nonunion or avascular necrosis in type II scaphoids, as well as potentially better success rates with antegrade screw placement rather than retrograde. Limited evidence currently exists on the clinical impact of different scaphoid types. Amundsen et al7 for example looked at eight patients with scaphoid proximal pole avascular necrosis, of whom 4 (50%) had type II scaphoids, supporting the theoretical increased risk in this group. A larger study focused on this issue is therefore needed to determine if scaphoid morphology impacts treatment outcomes. Our study did not include pathologic scaphoid x-rays to prevent skewing the potential results; therefore, there are no direct conclusions from this study that can be drawn regarding the impact of Morsy type on outcomes and complications (malunion, nonunion, avascular necrosis, etc.). However, there are potential clinical implications that can be made based on this knowledge. Surgeon awareness of morphological variation may impact management. Trikha et al8 analyzed scaphoid morphology in 50 healthy subjects and noted significant morphological variation that could impact surgical treatment. Achieving central screw placement is an established goal in the surgical management of scaphoid fractures, which could be impacted by scaphoid morphology.9,10 Although several studies have shown superiority in attaining the central axis of fixation using dorsal/antegrade screw placement compared with volar/retrograde screw placement,11,12 a volar approach can be more beneficial for comminuted fractures and/or nonunion reconstructions. Additionally, a large meta-analysis by Kang et al13 demonstrated no significant differences in complication rates, functional outcomes, nonunion rates, pain, or range of motion between dorsal and volar approaches. There are several limitations to this study. First, the retrospective nature limits the ability to determine causation regarding scaphoid type and demographics or clinical outcome. Additionally, the study subjects all had clinical indications for wrist imaging, which could compromise the external validity of our findings. The presence of wrist pathology may have also impacted the measurement values, despite attempts to minimize this by excluding those with inadequate imaging. Lastly, as this was not a prospective study, there was no ability to follow these patients over prolonged periods to see if any developed scaphoid pathology (eg, fracture, nonunion, Preiser disease, etc.). This is a potential area for future research. In conclusion, our work builds upon that of Morsy et al’s prior study to confirm a distinct difference in scaphoid morphology, with an approximately 20% prevalence of type II scaphoids in the general population. Despite the above limitations, we found no clinically relevant difference in comorbidities when comparing type I and II scaphoids. Future research should be aimed at elucidating potential differences in scaphoid fracture treatment outcomes between type I and II scaphoids, as well as longitudinal follow-up over a longer period in a larger cohort. Statement of informed consent The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of OHSU (no. 00025272) on January 3, 2023, with the need for written informed consent waived. Conflicts of Interest No benefits in any form have been received or will be received related directly to this article.