Prevalence of Ulnar Artery Thrombosis in Orthopedic Surgeons

• Updated on: 2019-06-19
• Read original article here

Ulnar artery thrombosis (UAT), or hypothenar hammer syndrome, occurs most commonly in manual laborers and high-performance athletes. The palm of the hand is repeatedly used as a substitute for a hammer or receives high impact from an object such as a racquet or jackhammer. Persistent and repeated impact damages blood vessels of the hand, specifically the ulnar artery as it passes through the Guyon canal. Damage to the ulnar artery in the hypothenar region as it passes over the hook of hamate results in a reduction of arterial blood flow to the fingers, frequently causing thrombosis and in some cases arterial aneurysm.

In a study of mechanical workers who habitually used their hands as “hammers,” Little and Ferguson reported a 14% (9 of 79) prevalence of ulnar artery occlusion compared with no reported cases in workers who did not use their hands in the same fashion. In addition, the authors surmised there was a correlation between ulnar artery damage and advancing age/longer employment. In a prospective study of 1300 participants enrolled during a 25-year period, Ferris et al determined the overall prevalence of hypothenar hammer syndrome was 1.6%, all of whom had occupational or avocational exposure to repetitive palmar trauma.

UAT symptoms include pain, paresthesia, sensitivity to cold, blanching, and calluses in the hypothenar region; the prevalence of these symptoms in orthopedic surgeons is unknown. The purpose of this study was to determine the prevalence of UAT in orthopedic surgeons and identify specific risk factors for developing UAT-associated symptoms.

This prospective study was undertaken to determine the prevalence of UAT and associated symptoms in orthopedic surgeons and residents. Participants were recruited at regional meetings and in practice settings. Eighty surgeons were recruited in the states of Arkansas and Maryland. Eligible participants included graduates from medical schools accredited by the Liaison Committee on Medical Education (LCME) or the American Osteopathic Association (AOA) as well as surgeons who either graduated from or were current residents in an accredited orthopedic surgery residency program. Participants were given an informational letter on the methods of testing, use of data obtained, and the reasoning behind the investigation. Institutional review board approval was obtained. Participants first completed a demographic questionnaire detailing their subspecialty training, years in practice, number of hip and knee arthroplasty cases performed monthly, and whether they had specific symptoms commonly associated with UAT. Symptoms of UAT included numbness of the hands and fingers, finger paresthesia, hand paresthesia, finger pain, hand pain, cold intolerance, and color change. In addition, participants were asked to report any leisure activities such as tennis, ping pong, golf, woodworking, weight lifting, and mechanical work that may increase their risk of UAT. A timed Allen test was performed by occluding the radial artery in all study participants, and time to reperfusion in seconds was recorded. Reperfusion was documented at the time when blanched areas of the hand returned to normal color. A reperfusion time greater than 6 seconds was classified as abnormal using the criterion established by Koman and Urbaniak, thereby constituting a positive result on the timed Allen test. Testing was performed by 3 of the authors (C.S.M., K.D., A.C.). Participants were considered to have UAT if they had UAT-specific symptoms accompanied by a positive Allen test. The Fisher exact test was used to compare the study UAT rate with historical UAT rates. The Fisher exact test also was used to compare participants with and without UAT for differences in binary participant characteristics, and to assess the strength of association between practice characteristics and the reported presence of specific UAT-related symptoms. In addition, analyses were performed with and without residents included. Wilcoxon rank-sum tests were used to compare participants with and without UAT for differences in length of practice and the number of hip and knee arthroplasties performed monthly. Wilcoxon rank-sum tests also were used to assess the strength of association between practice characteristics and the number of specific UAT-related symptoms reported. All statistical tests were 2-sided. Numeric P values were reported and interpreted heuristically for statistical significance using the sliding-scale approach of Mendenhall et al. Results were considered highly significant for P<.01, statistically significant for .01≤P<.05, trending toward statistical significance for .05≤P<.10, and not statistically significant for P≥.10. Statistical analyses were performed using SAS version 9.4 software (SAS Institute, Cary, North Carolina).

A total of 80 participants were enrolled in the study (Table 1). Of these, 25 participants were residents, and the remaining 55 participants were board-eligible or board-certified orthopedic surgeons. Seventy-one percent of the surgeons (39 of 55) reported their practice primarily comprised an orthopedic subspecialty. Five participants were subspecialty trained in hip and knee arthroplasty. The median (quartile) number of years in practice among participants was 10 years (range, 4–25 years) with residents included and 19 years (range, 10–27 years) with residents excluded. Of the 80 participants, 45 (56%) reported UAT-related symptoms in the hand. Nine (20%) of the 45 symptomatic participants and 1 (3%) of the 35 asymptomatic participants had a positive timed Allen test, for an overall UAT rate of 11% (9 of 45). This 11% rate among orthopedic surgeons was comparable (P=.64) to the 14% UAT rate reported by Little and Ferguson among 79 habitually hammering machinists but significantly higher (P<.0001) than the 1.6% UAT rate reported by Ferris et al among 1300 patients with hand ischemia. Table 1 lists the practice characteristics of study participants with and without UAT. All 9 UAT cases were present in physicians who had already graduated from an orthopedic surgery residency (P=.051). The number of years in clinical practice was a significant risk factor for developing UAT; 24% (8 of 33) of surgeons with 15 years or more of clinical practice had UAT compared with only 2% (1 of 47) of surgeons with less than 15 years of practice (relative risk=11.4; P=.0030). Furthermore, the median length of practice was 25 years for surgeons with UAT and 9 years for surgeons without UAT (P=.0054). There was no significant difference in the number of hip and knee arthroplasty cases performed by surgeons with and without UAT (Table 1). With residents excluded from the data analysis, the median length of practice was 9 years longer in the UAT group than in the group without UAT (P=.11). The UAT rates were 24% (8 of 33) among surgeons with 15 years or more of practice vs 5% (1 of 22) among surgeons with less than 15 years of practice (relative risk=5.33; P=.070). The UAT association with geographic region and the number of arthroplasties performed monthly remained nonsignificant with residents excluded (Table 2). Table 3 and Table 4 show the subspecialties of study participants by UAT group when residents were included and excluded. With residents included, the prevalence of UAT reached 40% (2 of 5) in adult reconstruction surgeons compared with only 9% (7 of 75) in the other subspecialties (relative risk=4.29), but the increase only trended toward significance (P=.095). No other subspecialty showed a significant association with UAT (Tables 3–4). Table 5 and Table 6 show the recreational activities of study participants by UAT group when residents were included and excluded. With residents excluded, all 9 UAT surgeons reported participation in at least 1 recreational activity (P=.096). Yoga was the only activity significantly associated with the presence of UAT (P=.032). Twenty-two percent (2 of 9) of the UAT-positive group reported participation in this activity. This result was likely due to the small population size, although in some practices of yoga, increased weight bearing on the hand may increase the risk of UAT. The average number of activities among participants with and without UAT averaged 1.78 and 1.20, respectively (P=.080). There was no significant correlation among participant characteristics and specific UAT-related symptoms including hand/digit numbness, finger paresthesia, cold intolerance, finger pain, or color change. In contrast, hand paresthesia showed highly significant associations with years of clinical practice and geographic region, and a statistically significant association with resident status (Table 7). Hand paresthesia was reported by 42% (14 of 33) of surgeons with 15 years or more of practice vs 4% (2 of 47) of surgeons with fewer years of practice (relative risk=9.97; P<.0001). Similarly, there was a 6.82-fold decrease in risk of hand paresthesia in residents (P=.016). The rate of reported hand pain reached 17% (8 of 48) in Arkansas participants vs 0% (0 of 32) in Maryland participants (P=.016). Surgeons with 15 years or more of practice had an 18% hand-pain rate compared with a 4% hand-pain rate among surgeons with less than 15 years of practice (relative risk=4.27), but the difference only trended toward statistical significance (P=.060). All 8 hand-pain reports came from surgeons who had graduated from residency, but this difference only trended toward statistical significance (P=.052). The number of UAT-related symptoms reported by participants was significantly affected by geographic region and years of practice, and also showed a trend toward significance with resident status (Table 7). The average symptom count was 1.60 for Arkansas surgeons vs 0.66 for Maryland surgeons (P=.022). In addition, the average symptom count was 1.85 for surgeons with 15 years or more of practice vs 0.79 for surgeons with less than 15 years of practice (P=.0016). The average symptom count was 1.44 among surgeons compared with 0.76 for residents (P=.096). Finally, the percentage of surgeons who reported at least 1 UAT-related symptom was 79% among those with 15 years or more of practice compared with 43% among those with less than 15 years of practice (P=.0014).

The prevalence of UAT in this cohort of orthopedic surgeons was 11%, which is significantly higher than the reported 1.6% among members of the general population. Excluding residents, the prevalence was 16%, which appears to be comparable with the prevalence observed in habitual hammerers. The increased prevalence in this cohort of orthopedic surgeons is possibly due to the amount of blunt force trauma to the hypothenar eminence of the hand, and symptoms increasingly correlated with advanced years in clinical practice. Although not statistically significant, there is a trend toward an increased prevalence in surgeons who specialize in hip and knee arthroplasty. Joint reconstruction may increase the risk of developing UAT because of the frequent use of vibrating instruments such as oscillating saws and power reamers, and the need to manipulate retractors and instruments forcefully. The heel of the hand sometimes is used as an impacting tool when placing Homan retractors about the hip or knee, or when inserting various implants. With the exception of yoga, no leisure activity was found to be significantly correlated with developing UAT in this population. The statistically significant association of yoga with UAT is interesting but is based on 2 UAT cases among 3 yoga practitioners, and further studies are needed to confirm this correlation. To be diagnosed with UAT, the current study required participants to have both UAT-related symptoms and a timed Allen test of 6 or more seconds. Although only 9 participants satisfied both criteria, 45 participants reported UAT-related symptoms. The timed Allen criterion of 6 or more seconds is drawn from the work of Koman and Urbaniak, but some authors argue a timed Allen test of greater than 2 seconds is often abnormal. Therefore, the criteria proposed by Koman and Urbaniak are more stringent and may have led to a lower rate of UAT diagnoses. If the criteria were loosened, many of the 36 symptomatic participants who did not meet the criterion of Koman and Urbaniak may have been diagnosed with UAT at the time of this study or are potentially at risk of developing it in the near future. Compared with Maryland participants, Arkansas participants had a significantly higher average number of UAT symptoms and significantly higher rates of both hand pain and hand paresthesia. Table 7 shows that for every single 1 of the 7 UAT-related symptoms shown, the Arkansas surgeons had a higher rate than the Maryland surgeons. It is possible that unknown geographic factors may lead to UAT, and therefore additional studies are necessary regarding this finding. This study has several limitations including the relatively small sample and limited geographic locations. Resident education and surgical training is often similar among physicians in 1 geographic area, thus some of this study population may be using particular surgical techniques that increase the risk of UAT development. Investigation among a larger group and a more diverse population is warranted. Definitive diagnosis of UAT must be confirmed with an angiogram, CT-angiogram, or MR-angiogram, which was not feasible for this study. Although obtaining a definitive diagnosis would make the data more accurate, it likely would not change the overall trend of the data results. Further investigation needs to be performed to determine whether hip and knee arthroplasty surgeons using this technique are truly at a greater risk than others. Were this the case, the traditional surgical techniques used for retractor and implant placement may need to be modified.