On "development of a clinical prediction rule ..." Currier et al. Phys Ther. 2007;87:1106-1119.

We read with interest the paper by Currier et al. (1) Although we do not dispute the authors' conclusion that their clinical prediction rule (CPR) can help identify candidates for hip mobilization in the population with knee pain, our alternative interpretation of their findings is that the CPR has helped to identify a subgroup of subjects with symptomatic hip osteoarthritis (OA) in their study cohort. Specifically, in the study by Currier and colleagues, all subjects except one had radiographic evidence or magnetic resonance imaging findings of hip OA. Given that the 2 CPR predictors of (1) hip or groin pain and (2) limited hip medial rotation are closely related to the clinical diagnosis of hip OA, (2) we think it is reasonable to conclude that most of the responders to hip mobilization were likely to have satisfied both the radiographic and clinical criteria of hip OA.

Identifying patients with symptomatic hip OA poses a diagnostic challenge. Khan et al (3) examined the pain location in the lower limb of 120 patients awaiting total hip arthroplasty or spinal decompression, and anterior thigh pain was found to be present in 70% of the patients with symptomatic hip OA. Furthermore, the researchers found that groin pain (positive likelihood ratio [PLR]=2.8) was the only region that distinguished symptomatic hip OA from a lumbosacral condition.

In 97 patients with lower-extremity pain, Brown and colleagues4 reported that groin pain and limited medial rotation of the hip were more closely associated with the presence of symptomatic hip OA than with the presence of spinal conditions. In a multicenter study of 195 patients with hip and groin pain, Birrell and colleagues (5) found that limited hip medial rotation (<23[degrees]) alone was most predictive of mild to moderate radiographic hip OA (PLR=2.5). In the study by Currier and colleagues, (1) among the CPR predictors, pain or paresthesia in the ipsilateral hip or groin had the highest PLR (8.10) in the bivariate analyses, whereas limited hip medial rotation (<17[degrees]) had the second highest PLR (6.02). Interestingly, when the standard error of measurement (6[degrees]) is considered in the interpretation of the latter predictor, the resultant cutoff value (upper bound) of hip medial rotation is remarkably similar to that found by Birrell and colleagues. (5)

In summary, we believe an equally tenable interpretation of the study by Currier et al is that subjects with radiographic and specific clinical symptoms of hip OA, as defined by the CPR, tended to respond favorably to hip mobilization.

Yong-Hao Pua, Boon Whatt Lim

YH Pua is doctoral candidate, Centre for Health, Exercise, and Sports Medicine, The University of Melbourne, Victoria, Australia. Address all correspondence to Yong-Hao Pua at: [e-mail].

This letter was posted as a Rapid Response on July 11, 2007, at www.ptjoumaLorg.

References

(1) Currier LL, Froehlich PJ, Carow SD, McAndrew RK. Development of a clinical prediction rule to identify patients with knee pain and clinical evidence of knee osteoarthritis who demonstrate a favorable short-term response to hip mobilization. Phys Ther. 2007;87:1106-1119.

(2) Altman R, Alarcon G, Appelrouth D, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum. 1991;34:505-514.

(3) Khan AM, McLoughlin E, Giannakas K, et al. Hip osteoarthritis: where is the pain? Ann R Coll Surg Engl. 2004;86:119-121.

(4) Brown MD, Gomez-Marin O, Brookfield KF, Li PS. Differential diagnosis of hip disease versus spine disease. Clin Orthop Relat Res. 2004;419:280-284.

(5) Birrell F, Croft P, Cooper C, et al. Predicting radiographic hip osteoarthritis from range of movement. Rheumatology (Oxford). 2001;40:506-512.

[DOI: 10.2522/ptj.2007.87.10.1407.1]

Author Response

We appreciate the response by Pua and Lim to our recently published article. (1) Although their interpretation that subjects in our study had concomitant hip OA is tenable, especially given the fact that up to 40% of subjects with knee OA have concomitant OA of the hip, (2) it is not the only explanation for our results. All 8 subjects in our study who had pain or paresthesia in the hip or groin on the same side as their knee pain had a successful response to hip mobilizations. In addition, 13 of 14 subjects who had passive hip medial rotation less than or equal to 17 degrees had a successful response to hip mobilizations. Three subjects (5%) had both pain or paresthesia in their hip or groin and hip medial rotation less than 17 degrees. However, only 2 of these 3 subjects met the criteria of Altman and colleagues (3) for a clinical diagnosis of hip OA with either test cluster 1 (hip pain, hip medial rotation less than 15[degrees], and hip flexion less than 115[degrees]) or test cluster 2 (painful hip medial rotation, greater than 50 years of age, and morning hip stiffness less than 60 minutes).

Although imaging abnormalities were present in all but one of our subjects in this study, many were considered mild. Although radiographic changes alone are not sufficient to establish the diagnosis of symptomatic hip OA, the involvement of surrounding soft tissues in these subjects may, in fact, be responsive to hip mobilization intervention. Many subjects who had a successful response to hip mobilizations did not meet any of the criteria or CPR variables correlated with clinical hip OA, yet they still responded successfully to the mobilizations. Indeed, 41 of the 60 subjects in our study had a successful response to hip mobilizations. This suggests that many subjects who do not meet Altman and colleagues' clinical criteria for hip OA also respond to hip mobilizations.

In our opinion, the most important finding of our study is that hip mobilizations appear to effectively relieve pain or improve patient status in a subgroup of patients with primary reports of knee OA. The precise pathoanatomic explanation at this time is unclear. Because hip symptoms are neither present nor pronounced in these patients, clinicians may overlook applying an intervention from which these patients may benefit. In summary, Pua and Lira's alternative interpretation explains only a subset of patients who responded successfully to the mobilizations. Patients who meet the other CPR variables cannot be discounted.

Linda L Currier, Robert S Wainner

References

(1) Currier LL, Froehlich PJ, Carow SD, et al. Development of a clinical prediction rule to identify patients with knee pain and clinical evidence of knee osteoarthritis who respond favorably to short-term hip mobilizations. Phys Ther. 2007;87:1106-1119.

(2) Aigner T, Dudhia J. Genomics of osteoarthritis. Curr Opin Rheumatol. 2003;15:634-640.

(3) Altman R, Alarcon G, Appelrouth D, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum. 1991;34:505-514.

[DOI: 10.2522/ptj.2007.87.10.1407.2]