The diagnostic validity of the cervical side bend-rotation test for C 1/2 dysfunction

ABSTRACT

Introduction

Neck pain and headaches are common, with a reported lifetime prevalence of up to 66%. Upper cervical segmental dysfunction has been implicated as meaningful in neck pain and multiple headache types. Several tests have been described to assess upper cervical joint dysfunction, including the flexion-rotation test (FRT), the side bend-rotation test (SBRT), and joint play assessment (PA). The purpose of this study was to determine the diagnostic validity of the SBRT to detect C1–2 dysfunction in a sample of people with medically diagnosed sinus headaches and controls.

Methods

Design: prospective diagnostic accuracy study, occurring during an observational case-control study in a sample of individuals with medically diagnosed sinus headaches. All participants were assessed using the SBRT, FRT, and C1–2 joint play assessments. The diagnostic accuracy of the SBRT was assessed using a reference standard of concurrent positive FRT (a loss of at least 10° from expected ROM (≤34°)) and restriction of C1–2 joint play. Cut-off scores for the SBRT were determined using ROC curve analysis, and tests of diagnostic accuracy were calculated using 2 × 2contingency tables.

Results

A total of 80 individuals (40 headache, 64 female, mean age 32.9 ± 13.8 yrs.) were included in the study. Mean ROM for the tests was: SBRT 31.4 ± 9.4°, FRT 44.9 ± 9.5°, and C1–2 mobility 22 hypomobile/58 normal. An SBRT cutoff score of <25° was confirmed using ROC curves. Using this cutoff score, the SBRT demonstrated 100% sensitivity and 62% specificity to detect C1–2 hypomobility.

Discussion/conclusion

The SBRT, using a cutoff score of ≤25°, appears to be a sensitive test to detect C1–2 dysfunction. Based on the strong sensitivity and negative predictive values, scores greater than 25° may effectively rule-out C1–2 dysfunction. The SBRT should be considered as part of a sequential clinical decision-making process when screening for C1–2 dysfunction, although further research is required to improve generalizability of these findings.

Introduction

Neck pain and headaches are prevalent across the general population, with 20–70% of the population experiencing neck pain at some point in their lives [1], and estimates of active headache disorders impacting 46% of the population with a lifetime prevalence of 66% [2]. Upper cervical segmental dysfunction, primarily of the C1-C2 segments, has been implicated as a significant factor in cervicogenic headache (CGH) [3–6]. However, this dysfunction is not confined to CGH, as similar dysfunction has been noted in other headache types including a 42% prevalence in headaches associated with rhinosinusitis (sinus headaches) [7], and a prevalence of 30–67% in individuals with migraine and mixed headache types [6,8]. There have been several tests described in the literature to assess upper cervical joint dysfunction [9], including the flexion-rotation test (FRT) [10], the side bend-rotation test (SBRT) [11], and upper cervical joint play assessment (PA) [5,12].

Perhaps most commonly used is the flexion-rotation test, which has been validated as a means to identify C1–2 dysfunction in a sample of individuals with cervicogenic headache, [5] although the test has also been used to identify C1–2 dysfunction in individuals with neck pain and other headache types as well [8,12]. By pre-positioning the spine in maximal flexion in supine, motion during subsequent craniocervical rotation is limited to primarily the upper cervical spine [13]. Hall and colleagues have reported the normal ROM to each side during the FRT to be 44°, with a suggested cutoff value of 34° to determine a positive test for individuals with cervicogenic headache [14]. This ROM estimate has most recently been supported by Zárate-Tejero et al, who reported median FRT ROM of 43.5–45.3º in a healthy population [9]. One challenge to isolating the C1–2 segment during the FRT is the reliance on maintenance of full lower cervical flexion, as well as the use of ligamentous tension to lock the lower cervical spine. When the FRT is performed actively by the patient, it is likely this ligamentous tension is not engaged, therefore the FRT is performed passively in order to detect side to side differences [15]. Takasaki et al reported that while the C1-C2 segment provides 73.5% of the total rotation during the FRT, the position reduces lower cervical ROM by a maximum of 76.9% [13]. This remaining ROM during the FRT may serve to mask subtle dysfunction, and may also be exacerbated by generalized hyperlaxity as progressive hyperlaxity has been demonstrated to result in greater rotational ROM [11].

An alternative test, the SBRT [9,11], which relies on osseous locking rather than soft tissue tension has been used clinically [16]. While coupled motions of the lower cervical spine (ipsilateral lateral flexion and rotation) have been described as inescapably linked, [17] the use of lower cervical non-coupled motions (contralateral side bending and rotation) create joint apposition and ligamentous tension, which has been described by multiple authors as a means to create locking/mechanical stability [16–19]. Mechanical coupling of the upper cervical spine (above C2) has been described as combined contralateral side bend-rotation motion; these mechanics allow motion to occur above C2 when the lower cervical spine is non-coupled or ‘locked’ [19–22]. Perhaps allowing for a more isolated C1–2 ROM than the FRT, the reported normal ROM is considerably less for the SBRT than the FRT, with previous studies of the SBRT reporting ROM norms between 31.25–37.7°[9,11]. This reported ROM aligns with studies performed with cadavers demonstrating mean C1–2 rotation of between 31.4–33.0º [23], and the 34.6–36.2º seen in vivo for C1–2 rotation by Kang et al and Ishii et al respectively [20,22]. Interestingly, the reported ROM for both the SBRT and C1–2 cadaver studies aligns with the portion of the FRT confirmed to C1–2 by Takasaki (44º*74% = 32.6º). [13] To date, the SBRT has not been validated in individuals with neck pain or headaches.

Joint play assessment has served as the gold-standard when assessing previous tests of upper cervical dysfunction, with C1–2 dysfunction established via joint play assessment serving as an inclusion criterion for the validation of the FRT for CGH. [5] While joint play assessment has been reported to have moderate-to-high reliability and strong sensitivity for assessment of pain [24,25], joint play assessment of the C1–2 segment has also been reported to be among the more reliable tests among those used to detect upper cervical dysfunction such as occurs with CGH [3]. In contrast, the overall assessment of mobility demonstrates only fair-to-moderate inter-rater reliability [25–27]. Research has shown that posterior-to-anterior mobilizations result in multi-segmental motion in a physiologic rather than solely accessory plane [28,29], and are influenced by a number of factors including local muscular response and practitioner estimations of stiffness. [30] Despite these limitations, there is otherwise a lack of a gold standard to determine spinal segmental movement dysfunction.

While Zárate-Tejero et al. recently compared ROM across tests in a healthy population, [9] the validity of the SBRT to detect upper cervical dysfunction in symptomatic populations remains unclear. Therefore, the purpose of this study was to determine the diagnostic validity of the SBRT to detect C1–2 dysfunction in a population of individuals with headaches and controls compared to the FRT and upper cervical joint play assessment.

Methods

This was a prospective diagnostic accuracy study, which took place as part of a larger observational case-control study assessing the prevalence of mechanical findings in a sample medically diagnosed with sinus headaches, the findings of which will be reported separately. Data were collected at a total of seven physical therapy clinics and university research labs located across the United States, with data collection occurring between February 2020 and February 2022. This study was approved by the Des Moines University institutional review board, IRB-2029-25 and prospectively registered at clinicaltrials.gov, NCT04222244. All participants provided written informed consent, and then attended one data collection session of approximately 30 minutes duration.

Investigators

All investigators conducting the cervical spine examination were licensed physical therapists with post-professional certification in manual therapy.

Participants

A convenience sample of individuals between the ages of 18–65 years were recruited via electronic advertising, word of mouth, and posted flyers seeking participants with a medical diagnosis of sinus headaches and headache free healthy age and gender matched controls in a 1:1 ratio. Individuals were excluded if they had a history of surgery to the cervical spine, whiplash injury within the past 5 years/persistent symptoms attributable to whiplash, and any headache types other than ‘sinus’ headaches.

Test and measures

All measurements were assessed randomly to minimize order effects by examiners blinded to the group allocation of the participants but not to the results of the index and reference tests.

Cervical flexion‐rotation test (FRT)

The FRT has been validated and shown to be reliable to identify C1‐2 dysfunction in patients with cervicogenic headache (CGH) and is currently the standard recommended measure for use in suspected cases of CGH [1,5]. With the participant positioned in supine, the participants’ neck was fully flexed by the examiner, then passively rotated left and right [5]. Cervical rotation ROM was measured using a digital goniometer [11,31]. Normal ROM has been estimated to be 44° [10], while a measurement loss of 10° or greater from expected ROM (≤34°) is considered a positive test result [14,32].

Side bend-rotation test

The side bend rotation test was performed as described by Swanson et al. [11] Participants were positioned supine, and their neck was then laterally flexed from C2 and progressing in a caudal direction. While lateral flexion was maintained, the participants’ head was then rotated in the contralateral direction by the examiner. This was repeated to the opposite side, and ROM measures were assessed using a digital goniometer mounted to a secure headpiece [11].

To determine a positive SBRT result, the expected mean ROM was calculated from prior studies [9,11], with a weighted mean ROM of 35°. Consistent with Hall et al, [14,32] a result was considered limited when the ROM was a minimum of 10° less than the expected ROM; [5] in other words, a positive test for C1–2 dysfunction was SBRT ROM of ≤25°. This estimate was analyzed during statistical analysis using a series of ROC curves.

Manual segmental joint examination

Passive segmental mobility of the upper cervical spine, specifically C1–2, as described by Maitland [33] was performed by the investigators. Participants were placed in the prone position, with the head rotated approximately 30° ipsilaterally, and a unilateral P-A force was applied to the articular pillar of C2 [12]. Motion was graded according to the quality of resistance on a 4‐point scale (normal, slight, moderate, marked resistance), and subsequently coded as hypomobile (moderate-marked) or not hypomobile (normal-slight). Previous research has demonstrated this method to be reliable for patients with neck pain and headaches, [34,35] and has been as the reference standard in C1–2 dysfunction in prior diagnostic validity studies. [14]

Reference standard

While no true ‘gold standard’ test exists for clinical upper cervical joint dysfunction, both the FRT and C1–2 joint play have demonstrated high levels of reliability and diagnostic validity. To determine the diagnostic accuracy of the SBRT to detect C1–2 dysfunction, concurrent findings of a positive FRT noting a loss of at least 10° from expected ROM (≤34°) [5,10,14] as well as restriction of C1–2 joint play [5] served as the reference standard.

Statistical methods

Sample size: Based on the work of Bujang and Adnan, [36] assuming a 50% sensitivity as the null hypothesis, and an alternative hypothesis of 80% or higher sensitivity, and a prevalence estimate of 30% to account for anticipated low occurrence rates within the control group (previous work identified a 42% prevalence of C1–2 dysfunction in a population self-identified as having sinus headaches [7], a minimum sample of 67 individuals would be required to achieve 80% power. As the true prevalence of C1–2 dysfunction in the medically diagnosed sinus headache population is unknown, and to account for potential missing data, the sample size was increased to 82.

Statistical analysis was performed using statistical software: SPSS (version 28, IBM, Chicago, IL) and MedCalc Software Ltd. diagnostic test evaluation calculator. (Version 22.020; accessed 20 February 2024). All data were analyzed quantitatively in aggregate form; the level of significance was established a priori with the α value set to 0.05 (type I error) and the β set at 0.2 (type 2 error). All data were assessed for extreme outliers (3rd quartile + 3*interquartile range; 1st quartile − 3*interquartile range), assessed for normality using the Shapiro-Wilk test, visual inspection of the Q-Q plot and the Levene statistic for homogeneity of variance using SPSS. Data were assessed both with and without outliers; as no significant differences were observed the original data were maintained for analysis. In the event of missing data, the data point was excluded from analysis.

Baseline data for groups were analyzed with descriptive statistics, and between group differences were assessed with independent samples t-tests for continuous variables, and the Χ² test of independence was used for nominal data. To better understand the relationships between the various tests as a measure of concurrent validity, correlation coefficients were calculated for SBRT and FRT ROM (Pearson’s r), and for SBRT and FRT ROM relative to C1–2 hypomobility (point biserial), with Perfect = 1.0, Strong = 0.7–0.9, Moderate = 0.4–0.6, weak = 0.1–0.3. [37] Prior to proceeding with calculation of diagnostic accuracy, ROC analysis and subsequent receiver operator characteristic (ROC) curves were calculated for the SBRT to verify the most appropriate cutoff score. ROC curves combine sensitivity and specificity with the area under the curve (AUC). The area with maximal AUC was determined to serve as the appropriate cutoff score for a positive SBRT result, maximizing sensitivity while maintaining a reasonable level of specificity. Diagnostic accuracy data was then analyzed based upon the dichotomous scoring of ‘positive’ or ‘negative’ for each test as previously described. Two by two contingency tables were analyzed using MedCalc to calculate sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV), and likelihood ratios with 95% confidence intervals for the SBRT relative to the reference standard of combined FRT and C1–2 joint play assessment, as well as the FRT and C1–2 joint play assessment individually. Finally, diagnostic odds ratios (DOR) were calculated to determine an overall measure of test performance that is independent of prevalence.[38]

Results

Complete data were available for a total of 80 participants, as data for two participants were incomplete and did not allow analysis. Participant demographics and baseline measures are included in Table 1. Participant flow is detailed in the STARD diagram. (Figure 1) There were no adverse events reported at any time during the study period.

Table 1.

Participant demographics and baseline values.

	Headache participants (n=40)	Control participants (n=40)	p value
Age (yrs.)	33.2 ± 14.2	32.5 ± 13.3
Weight (Kg)	75.3 ± 18.7	73.4 ± 16.9
Sex (%female)	32 (80%)	32 (80%)
SBRT ROM	Overall Mean ROM: 31.4 ± 9.4º
	Left: 28.4 ± 8.2° Right: 28.7 ± 7.2°	Left: 35.1 ± 11.4° Right: 33.2 ± 8.7°	.003 .013
FRT ROM	Overall Mean ROM: 44.9 ± 9.5
	Left: 42.9 ± 8.7° Right: 45.11 ± 8.8°	Left: 45.6 ± 10.2° Right: 45.9 ± 10.1°	.216 .717
C1-2 Restriction	23 no/17 yes	35 no/5 yes	.003

SBRT = side bending rotation test, ROM = range of motion, FRT = flexion rotation test; Significance p ≤ .05.

Figure 1.

STARD diagram.

Assessing the concurrent validity of the measures, overall correlations between the SBRT and FRT for ROM were moderate, right r=.440 [.243, .603], left r=.508 [.325, .655]. [39] Restricted C1–2 mobility was weakly correlated with SBRT ROM, right r_pb = −0.191, left r_pb = −0.275 and FRT ROM, right r_pb = −0.268, left r_pb = −0.206.

A series of ROC curves were analyzed to determine the maximum area under the curve to confirm the cutoff score of SBRT. The resulting AUC results were SBRT ≤ 23° = 0.608, SPRT ≤ 24° = 0.588, SBRT ≤ 25° = 0.811, and SBRT ≤ 26° = 0.791. (Figure 2).

Figure 2.

ROC Curve, degrees of ROM during the SBRT.POS=positive.

When analyzed for diagnostic accuracy against the reference standard of combined (+) FRT ≤ 34º and restricted C1–2 joint play, the SBRT demonstrated a sensitivity of 100% [54.1, 100], specificity of 62.1% [50.1, 73.2], a positive likelihood ratio (LR+) of 2.64 [1.97, 3.54], and a negative likelihood ratio (LR-) of 0. The resulting positive predictive value (PPV) was 17.7% [13.8, 22.3], while the negative predictive value (NPV) was 100% [92.3, 100] based on an overall diagnostic accuracy of 65% and an observed prevalence of 7.5% (Table 2).

Table 2.

2×2 contingency table: SBRT vs FRT and C1–2 joint play assessment.

Test	FRT + C1–2 +	FRT- C1–2 -	Total
SBRT +	6	28	34
SBRT -	0	46	46
Total	6	74	80

SBRT= side bending rotation test, FRT = flexion rotation test, FRT + = positive reference test, FRT - = negative reference test, C1–2 + = positive reference test, C1–2 - = negative reference test.

When comparing the accuracy of the SBRT against the reference standard of the FRT, the SBRT demonstrated a sensitivity of 68.8% [41.3, 89.0], specificity of 64.1% [51.1, 75.7], LR + 1.91 [1.20, 3.05], LR- 0.49 [0.23, 1.03], PPV 32.4% [23.1, 43.2], NPV 89.1 [79.5, 94.6] base on an overall diagnostic accuracy of 65% and a prevalence of 20% (Table 3).

Table 3.

2×2 contingency table: SBRT vs FRT.

Test	FRT +	FRT -	Total
SBRT +	11	23	34
SBRT -	5	41	46
Total	16	64	80

SBRT= side bending rotation test, FRT = flexion rotation test FRT + = positive reference test, FRT - = negative reference test.

When comparing the accuracy of the SBRT against a reference standard of C1–2 joint play, the SBRT demonstrated a sensitivity of 59.1% [36.4, 79.3], specificity of 63.8% [50.1, 76.0], LR + 1.63 [1.00, 2.66], LR-0.64 [0.37,1.10], PPV 38.24% [27.6,50.2], NPV 80.4 [70.6, 87.6] based on an overall diagnostic accuracy of 62.5% and a prevalence of 27.5% (Table 4).

Table 4.

2×2 contingency table: SBRT vs C1–2 joint play assessment.

Test	C1-2 +	C1-2 -	Total
SBRT +	13	21	34
SBRT -	9	37	46
Total	22	58	80

SBRT= side bending rotation test, C1–2 + = positive reference test, C1–2 - = negative reference test.

For comparison, when assessing the accuracy of the FRT to a reference standard of C1–2 joint play assessment, the FRT demonstrated sensitivity of 27.7% [10.7,50.2], specificity of 82.8% [70.6,91.4], LR + 1.58 [0.65,3.83], LR- 0.88 [0.66,1.6], PPV 37.5% [19.8,59.3], NPV 75% [69.4,79.9] based on an overall diagnostic accuracy of 67.5% and a prevalence of 27.5% (Table 5).

Table 5.

2×2 contingency table: FRT vs C1–2 joint play assessment.

Test	C1-2 +	C1-2 -	Total
FRT +	6	10	16
FRT -	16	48	64
Total	22	58	80
FRT = flexion rotation test.

A summary of all test results is provided in Table 6.

Table 6.

Summary of diagnostic accuracy across tests.

	SBRT x FRT and C1–2	SBRT x FRT	SBRT x C1–2	FRT x C1–2
Sensitivity	100	68.8	59.1	27.7
Specificity	62.1	64.1	63.8	82.8
LR+	2.64	1.91	1.63	1.58
LR-	0	0.49	0.64	0.88
PPV	17	32.4	38.24	37.5
NPV	100	89.1	80.4	75.0
DOR	21.2 [1.15, 390.93] *	3.92 [1.21, 12.68]	2.55 [0.93, 6.95]	1.80 [0.56, 5.74]

LR+ = positive likelihood ratio, LR- = negative likelihood ratio, PPV = positive predictive value, NPV = negative predictive value, DOR = Diagnostic odds ratio, * = 0.5 added to all cells (a, b, c, d).

Discussion

A finding of SBRT ≤ 25º, when compared to a reference standard of ≤34º FRT and positive C1–2 joint play findings, demonstrated 100% sensitivity to detect C1–2 dysfunction within this sample of 80 individuals with sinus headaches and controls. The resulting NPV of 100% indicates that a SBRT test result of greater than 25º will essentially rule out the presence of C1–2 dysfunction. Some caution is warranted in the interpretation of this definitive NPV, as the low prevalence of individuals meeting the reference standard of both restricted C1–2 joint play and positive FRT within the sample may have resulted in an artificially low NPV, as NPV decreases with an increasing prevalence of the disease within a sample. [38] The overall prevalence of combined positive FRT and C1–2 joint dysfunction (P-A) was much lower than anticipated at only 7.5%. The C1–2 joint dysfunction of 27.5% was in line with the anticipated sample, while the FRT prevalence of 20% was lower than anticipated. Interestingly, there was not substantial overlap between the two tests, as only 6/22 individuals with positive C1–2 restriction during P-A assessment also presented with a positive FRT, and the FRT demonstrated acceptable specificity but low sensitivity to detect C1–2 dysfunction.

Considering the concurrent validity of the index and reference tests, not surprisingly there was only a moderate correlation between the SBRT and FRT for overall ROM. The mean ROM findings for both tests were similar to those observed in previous studies which demonstrated significant differences between FRT and SBRT ROM in asymptomatic individuals [9,11], and the SBRT ROM in this study was similar to isolated C1–2 motion demonstrated in cadaver models. [23] There were statistically significant differences in SBRT ROM and the presence of/absence of C1–2 joint mobility restrictions (P-A) between individuals with headaches and controls, which supports the validity of the SBRT as assessing this construct. Conversely, there were no differences noted for the FRT between groups. There are several possible explanations for this difference. Previous work has shown that the FRT is more susceptible to the influences of hypermobility than the SBRT. [11] The fully flexed position relies on soft-tissue tension and has been shown to allow approximately 74% of the rotational motion during the FRT to occur at C1–2, with the remaining motion occurring in other cervical segments. [13] In light of this reliance on soft tissue tension it is also plausible that, since the FRT is a test of rotation in maximum cervical flexion, prone C1–2 passive segmental joint play assessment (relatively extended posture) may not stress either the surrounding soft tissues or articular structures in the same way. Despite these differences, it is important to remember that both the FRT and SBRT were only weakly correlated with SBRT mobility. During the validation study by Ogince et al investigating the FRT within a population with CGH, a preliminary screening for symptomatic C1–2 dysfunction was included prior to progressing to the FRT phase, [5] thus the diagnostic accuracy of the FRT to detect C1–2 hypomobility (as it is often used clinically) [12] may not match the reported accuracy for CGH as PPV increases with increased prevalence. [40] These findings do call into doubt the use of the FRT as a screening procedure for C1–2 dysfunction as opposed to use as a screening procedure for symptomatic CGH. This finding for the FRT is in agreement with the findings of Rubio-Ochoa et al. and the 2017 Neck Pain CPG, which suggest that the FRT may be most useful at the later stages of examination as a confirmatory technique. [1,3] Conversely, it appears that the SBRT has greater utility to screen for C1–2 dysfunction at earlier stages of the examination.

The trade-off of the sensitive SBRT cutoff score is a tendency toward a high number of false positive SBRT findings. These false positives may need additional consideration from a mechanical viewpoint. Not all authors have agreed upon the coupling behavior of the upper cervical spine, with greater variability noted if side bending is the initial motion compared to initiation of rotation, with upper cervical side bending and rotation coupled in the same direction. [16] While the SBRT initiates side bending from C2,[11] in part to avoid any additional soft tissue stresses or impact on coupling mechanics of C1–2, it is not clear if this has a similar effect on coupling mechanics as side bending initiated from the cranium. Additionally, coupled contralateral side bending and rotation in the C1–2 segment is reported to be accompanied by a small amount of flexion. [22] As the SBRT is generally performed with the participants head in contact with the plinth, the test position may have created upper cervical extension and thus limited the available coupled motion, particularly in cases where the participant had a greater degree of thoracic kyphosis. In an exploratory analysis, there were five participants with SBRT < 25° bilaterally and rated as normal C1–2 joint play; among these five individuals the mean SBRT was 19.3°, while the mean FRT was 41.0°. If both measures were equally limited, it would be likely that the finding was simply the result of the well acknowledged wide variation of upper cervical segmental motion frequently encountered in asymptomatic populations. [41] The discordant findings in this subgroup appear to support differences in coupling mechanics of the upper cervical spine in a small percentage of individuals, although it is not clear from the available data whether this is the result of anatomic variation or the presence of upper cervical extension during the test set-up. While speculative, it may be possible to identify this group- or potentially tester error- by the presence of bilaterally restricted SBRT ROM in the presence of normal FRT and joint play assessments.

There are several limitations to consider for this study. The low prevalence of positive findings within the sample is a considerable limitation of this study. Despite thoughtful design including a conservative sample estimate, the study sample did not have the same prevalence for the reference tests that was observed previously, and this study is therefore underpowered. This is evident in the very wide confidence intervals observed in the results. Methodologically, this diagnostic accuracy study was housed within the context of a case-control study assessing individuals with sinus headache compared to controls. While assessors were blinded regarding the group allocation, they were not blinded regarding the index and reference tests in this portion of the study. While the researchers feel the overall design protects the integrity of the data, it is possible that the lack of blinding introduced a source of bias. Reference standards were selected based on the existing evidence. The overall lack of correlation between the two selected reference tests (FRT and C1–2 joint play), despite their application by examiners with considerable experience with the techniques, lends some doubt to the utility of either test to accurately assess the mobility of the C1–2 segment. It is important to note, however, that the two tests assess motion under very different conditions and may not be equivalent constructs. Finally, the population of individuals with sinus headache, despite previous research demonstrating a significant prevalence of upper cervical dysfunction, [7] may not be representative of the general population of individuals with neck pain thus limiting the generalizability of findings.

Conclusions

The SBRT appears to be a sensitive test which assists in the identification of C1–2 segmental dysfunction when using a cutoff score of ≤25º Based on the strong sensitivity and negative predictive values, scores greater than 25º may effectively rule-out C1–2 dysfunction. The SBRT should be considered as part of a sequential clinical decision-making process when screening for C1–2 dysfunction, although further research is required to improve generalizability of these findings.

Biographies

Brian T. Swanson, PT, DSc is an Associate Professor at the University of Hartford. He serves as Chair of the Department of Rehabilitation Sciences, Director of the Doctor of Physical Therapy program, and co-director of the University of Hartford/HHCRN orthopedic physical therapy residency program. He is board certified in orthopaedics and a Fellow of the American Academy of Orthopaedic Manual Physical Therapists. Dr. Swanson’s research interests include validating tests and measures in orthopedic manual physical therapy, developing a further understanding of the mechanisms of manual physical therapy interventions, and evidence-based practice/research methodology.

Kenneth E. Learman, PT, PhD, FAAOMPT is a Distinguished Professor in the Department of Graduate Studies in Health and Rehabilitation Sciences at Youngstown State University. He is the director for the Doctorate in Health Sciences along with teaching EBP and orthopedic manual therapy in the DPT program. With 35 years of OMT experience, he is a senior deputy editor for the Journal of Manual and Manipulative Therapy and has served as a reviewer for more than a dozen other journals. He has published over 60 manuscripts, 7 book chapters and co-edited an orthopedic cases textbook.

Shannon M. Petersen, PT, DScPT, OCS Emeritus, FAAOMPT is a Professor in the Doctor of Physical Therapy Program at Des Moines University in West Des Moines, Iowa, teaching primarily musculoskeletal spine and manual therapy content. She is a Fellow of the AAOMPT and has been recognized as an OCS Emeritus by the American Physical Therapy Association. Dr. Petersen’s research interests relate to the cervical spine and headaches. She serves as a reviewer for multiple professional journals and conferences.

Bryan O’Halloran is an Associate Professor at the Medical University of South Carolina in the Division of Physical Therapy. He is a Fellowship trained clinician with specialties in Orthopedics and Sports, a subspecialty in Orthopedic Manual Physical Therapy and additional formal training (DSc(PT)) in clinical research methods. He has over 3 decades of clinical experience and has devoted the last 12 years to expanding clinical research and understanding the conservative management and mechanisms of manual therapy primarily associated with neck pain, back pain and shoulder conditions.

Funding Statement

Shannon Petersen, Kenneth Learman, and Bryan O’Halloran received funding from the Orthopedic Physical Therapy Products Grant through the American Academy of Orthopaedic Manual Physical Therapists. Brian Swanson received funding through the Dean’s Office, College of Education, Nursing, and Health Professions, University of Hartford.

Disclosure statement

No potential conflict of interest was reported by the author(s).