Detection of the palmaris longus tendon: physical examination versus sonography

Abstract

This study investigated the sensitivity, specificity, and accuracy of physical examination compared with high resolution sonography for the detection of the palmaris longus tendon in 136 wrists. The incidence of tendon absence was 10% by sonography and 14% by physical examination. The sensitivity of the physical examination was 94% and the specificity 86% compared with sonography as the reference standard. Accurate identification of the palmaris longus tendon through physical examination was not influenced by body mass index or wrist circumference. Upon review of the sonography images, none of the palmaris longus tendons missed on physical examination were considered to be adequate for grafting based on expert opinion. While physical exam was 93% accurate for detection of the palmaris longus tendon compared with ultrasound, our study suggests that it is sufficient for detection of those tendons that can be used as donors clinically.

Introduction

The palmaris longus (PL) tendon is one of the most commonly used donors for autologous tendon graft or tendon transfer (Cain et al., 2010; Jakubietz et al., 2011; Kocaoglu et al., 2017). Confirmation of the presence of a PL tendon therefore, is critical for planning any of these procedures. Physical examination maneuvers have been described that facilitate identification of the tendon (Kigera, 2012; Kigera and Mukwaya, 2011; Nasiri et al., 2016).

The PL tendon is estimated to be congenitally absent in approximately 15% of the population. However, the reported incidence of absence varies considerably by ethnicity, from 2% in Zimbabwe (Gangata, 2009) to 64% in Turkey (Ceyhan and Mavt, 1997), and by detection method (Andring et al., 2018; Ioannis et al., 2015; Kyung et al., 2012). In addition to congenital absence of the PL, anomalous variations have been described (Andring et al., 2018). The tendon may be duplicated, reversed, bifid, or divided to form an ulnar accessory slip distally. Described anomalous origins of the muscle belly include the flexor digitorum superficialis, flexor carpi radialis, flexor carpi ulnaris, and lacertus fibrosus. Anomalous insertions include the flexor carpi ulnaris, antebrachial fascia, thenar eminence, transverse carpal ligament, or carpal bones. A meta-analysis published by Pekala et al. (2017) indicated that functional studies tend to overestimate the absence of the PL tendon when compared with cadaveric studies. Additionally, detection of the PL tendon by physical examination may become less accurate in obese patients. Sonography provides accurate assessment of the PL tendon for both its presence (100% sensitivity and specificity) and measured dimensions in cadavers (Grechenig et al., 1999; Johnson et al., 2018).

The primary goal of our study was to establish the sensitivity, specificity, and accuracy of the physical examination compared with sonography in the detection of the PL tendon. Additionally, we hypothesized that physical examination is less accurate in patients with increased body mass index (BMI) and wrist circumference. Lastly, in an attempt to study the importance of experience in the accuracy of physical examination, we hypothesized that a fellowship-trained hand surgeon would be able to more accurately identify a PL tendon on physical examination than a fourth-year medical student.

Methods

Patient selection

The study evaluated 68 patients selected from a single upper extremity orthopaedic clinic. Participants were of mixed ethnicity from an urban centre in the southeastern United States. Exclusion criteria included patients younger than 18 years old, those unable to participate in examination (such as those in a splint covering the wrist), or those recovering from hand or wrist surgery. On consecutive days, potential patients were approached for participation. After completion of data collection and sonographic evaluation for a participant, the next available patient by appointment time order who met inclusion criteria was approached for participation.

Physical examination

After informed consent in compliance with institutional policy, patients underwent physical examination of both wrists by two separate examiners: a hand-fellowship-trained orthopaedic surgeon and one of two fourth-year medical students. Prior to initiation of the study, each medical student was given a demonstration of the physical examination techniques for identification of the PL tendon. Each examiner was blinded to the results of any prior examination. Physical examination was carried out using four well-described physical examination manoeuvres designed to facilitate identification of the PL tendon, including Schaeffer’s (or Standard), the Pushpakumar’s, Thompson’s, and Mishra’s tests (Figure 1) (Kigera and Mukwaya, 2011; Nasiri et al., 2016). The patient’s sex, BMI, and wrist circumferences were recorded.

Figure 1.

Physical exam manoeuvres. (a) Schaeffer’s standard test – opposition of thumb to small finger with slight wrist flexion. (b) Pushpakumar’s test – extension of the index and middle fingers, flexion of the ring and small fingers with thumb opposed and flexed. (c) Thompson’s test – a fist is formed by opposing and flexing the thumb over the flexed fingers, followed by resisted wrist flexion. (d) Mishra’s second test – resisted thumb abduction with partial wrist flexion.

Sonographic examination

Following completion of physical examination by both medical student and hand surgeon, each patient underwent sonographic examination by a sports medicine fellowship-trained primary care physician with a Registered in Musculoskeletal Sonography certification and who was blinded to the results of physical examination. High frequency sonography was used to determine the presence or absence of the PL tendon in each wrist. In cases in which physical examination and sonography were discordant, the sonographic images were reviewed to determine potential causes of error in examination (e.g. aberrant PL insertion, or relatively small tendon). An evaluation of the fitness of the tendons for grafting was also done. Since no well defined or sonographic parameters for the suitability of a PL tendon as a graft have been established or clinically validated, the evaluation was purely subjective, but consensus between a group of hand surgeons was established.

Statistical methods

Sonography served as the reference standard when calculating sensitivity, specificity, accuracy, and predictive values for physical examination detection of the PL tendon. Statistical analysis was done on subgroups to assess the impact of BMI, wrist circumference, sex, and examiner experience on the incidence of errors and accuracy of physical examination. Student’s t-tests was used for comparing the BMI and wrist circumference for the wrists with errors vs. the wrists without errors. Fisher’s exact test was used for comparing the incidence of error between subgroups of patients. A p-value < 0.05 was considered statistically significant.

Results

Detection rate by physical examination versus sonography

Sixty-eight patients with 136 wrists were included. Thirty-one patients were men, 37 women. Mean BMI was 28 kg/m² (SD 6.4). Mean wrist circumference was 17 cm (SD 1.5)

The incidence of PL tendon absence by sonography and by physical examination is listed in Table 1. The fellowship-trained hand surgeon found the PL tendon to be present by physical examination in 117 wrists, two of which were found to be false positives (FP) by sonography. The surgeon reported the tendon absent in 19 wrists, seven of which were false negatives. Overall, physical examination was 93% accurate for detection of the PL tendon compared with sonographic evaluation. The sensitivity of the physical examination compared with sonography as the reference standard was 94%, and the specificity was 86%. The predictive value of a positive finding was 98% and of a negative finding 63%.

Table 1.

Incidence of palmaris longus tendon absence by physical examination and sonography in 68 patients (136 wrists).

	Overall incidence of absence in wrists	%	Incidence of unilateral absence in patients	%	Incidence of bilateral absence in patients	%
Physical examination
By hand surgeon	19/136	14
By medical student	13/136	10
Sonography
Total sample	14/136	10	6/68	9	4/68	6
Men	8/62	13	4/31	13	2/31	6
Women	6/74	8	2/37	5	2/37	5
Right wrist	6/68	9	—		—	—
Left wrist	8/68	12	—		—	—

There were no statistically significant differences in error of detection by physical examination between the subgroups defined by BMI, wrist circumference, sex, and examiner experience (Table 2). In two patients with BMI ≥50, the PL tendon of all four wrists was correctly identified with physical examination. The mean BMI of the wrists with errors on physical examination was 29 kg/m² (SD 5.4) vs. 28 kg/m² (SD 6.4) in the wrists without errors (p = 0.74). The mean wrist circumference of the wrists with errors was 17 cm (SD 1.4) compared with 17 cm (SD 1.5) in the wrists correctly identified (p = 0.90).

Table 2.

Rate of error of physical examination by body mass index, wrist circumference, sex, and examiner.

	Rate of errora	%	p-valuec
Body mass index
>30 kg/m2	3/44b	7	1.00
≤30 kg/m2	6/92b	7
Wrist circumference
>17 cm	3/53b	6	1.00
≤17 cm	6/83b	7
Sex
Men	3/62b	5	0.51
Women	6/74b	8
Examiner
Hand surgeon	9/136	7	1.00
Medical student	9/136	7

^aNumber of false positives + number of false negatives.

^bPhysical examination done by hand surgeon.

^cFisher’s exact test.

Characteristics of palmaris tendons not being detected by physical examination

The fourth-year medical student made nine errors in detection of the PL tendon on physical examination compared with sonography, the same number of errors as the fellowship-trained hand surgeon. This included five false positives and four false negatives. The hand surgeon and medical student made the same error in four wrists. Each made five additional unique errors. Upon review of the false positive cases, we noted that the medical students were frequently misidentifying the flexor digitorum superficialis tendon to the middle finger as the PL tendon.

After review of the sonographic images, there was consensus among the hand surgeons that the seven PL tendons missed on physical examination were likely inadequate for grafting and that in clinical practice the surgeons would opt for an alternative graft material.

Discussion

In our study of 136 wrists, we documented an absence of the PL tendon in 14% of wrists by physical examination and 10% of wrists by sonography. These incidences are consistent with a meta-analysis of previously reported absences based on 49 physical examination studies (16% absences) and 19 cadaveric studies (10% absences) (Pekala et al., 2017). Overall, physical examination by a fellowship-trained hand surgeon was accurate in detection of the PL tendon compared with sonographic evaluation with 94% sensitivity, 86% specificity, and 93% accuracy. However, two false positives and seven false negatives occurred with physical examination.

An important consideration when harvesting the PL tendon, and an important finding of our study, was the low false positive incidence on physical examination. In the clinical setting, the false positive incidence is potentially the more dangerous error as it could lead to attempted harvest of the presumed present PL tendon and result in inadvertent harvest of a more functionally important tendon or, disastrously, the median nerve (Leslie et al., 2017; Woon et al., 2017).

False negatives occurred at a higher incidence but are the less serious of the two errors in the clinical setting. All seven PL tendons missed on examination were small or wispy structures with anomalous or broad, indefinite insertions. Upon review of these seven cases, none of these tendons were considered to be suitable for tendon grafting by the fellowship-trained hand surgeon and senior author. Harvey suggested that only those tendons with a minimum diameter of 3mm were useful as grafts (Harvey et al., 1983). In a cadaveric study, Jakubietz et al. (2011) concluded that the majority of PL tendons were of adequate diameter (using a cutoff of 3mm) but that 63% did not meet his proposed requirement for a length of 15 cm. However, no strict anatomic- or sonography -based parameters for the suitability of a PL tendon to be used as a graft have been established or clinically validated and therefore the evaluation of suitability in our study is based merely on subjective expert opinion. Furthermore, given its accuracy in determining the diameter and length of the PL tendon, sonography may be useful to surgeons planning a procedure that requires a specific length or diameter (Johnson et al., 2018).

We hypothesized that physical examination would be less accurate in patients with increased BMI and wrist circumference; but based on our findings, we rejected the hypothesis. Furthermore, because physical examination was equally effective regardless of BMI or wrist circumference, there appears to be no additional benefit of sonography in obese individuals for detection of the PL tendon.

Additionally, both a fourth-year medical student and a fellowship-trained hand surgeon made nine errors each. However, the medical student had a higher incidence of misidentification of the adjacent flexor digitorum superficialis tendon to the middle finger as being the PL tendon (false positives). Thus, we suggest an additional technique to facilitate differentiation between these two tendons by palpating for excursion during middle finger flexion or palpating for tendon tensing during resisted middle finger flexion: The PL tendon should have no palpable excursion and no tensing.