Critical appraisal checklist for Randomized Controlled Trials (RCT)

Adapted from JBI WEBSITE: HTTPS://JOANNABRIGGS.ORG/ -CRITICAL APPRAISAL

TOOLS: HTTPS://JOANNABRIGGS.ORG/CRITICAL-APPRAISAL-TOOLS

ARTICLE TITLE:
Hip arthroscopy versus physical therapy for the treatment of symptomatic acetabular labral tears in
patients older than 40 years: a randomized controlled trial.

Martin SD, Abraham PF, Varady NH, et al. Hip arthroscopy versus physical therapy for the
treatment of symptomatic acetabular labral tears in patients older than 40 years: a randomized
controlled trial. Am J Sports Med. 2021;49(5):1199-1208. doi:10.1177/0363546521990789.

Reviewer: Isaac Weir

Yes No Unclear NA
1. Was true randomization used for assignment of X
participants to treatment groups?
2. Was allocation to treatment groups concealed? X

3. Were treatment groups similar at the baseline? X

4. Were participants blind to treatment assignment? X

5. Were those delivering treatment blind to treatment X

assignment?
6. Were outcomes assessors blind to treatment assignment? X

7. Were treatment groups treated identically other than the X

intervention of interest?
8. Was follow up complete and if not, were differences X
between groups in terms of their follow up adequately
9. described and analyzed?
Were participants analyzed in the groups to which they X
were randomized?
10. Were outcomes measured in the same way for treatment X
groups?
11. Were the instruments used to measure outcomes reliable X
and valid?
12. Was appropriate statistical analysis used? X

13. Was the trial design appropriate, and any deviations from X
the standard RCT design (individual randomization,
parallel groups) accounted for in the conduct and analysis
of the trial?

© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.
Overall credibility of article results per your assessment on the scale of 0-10, with 0 –“I don’t trust
the results, as the intervention study outcomes are questionable” to 10 – “I will definitely use the
results of the study in planning interventions for my patients” _______________8____________-

© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.
Explanation of critical appraisal of Randomized Controlled Trial (RCT)

Abstract:
This unblinded randomized controlled trial was intended to answer the question of the
effectiveness of physical therapy alone or physical therapy for post-operative hip acetabular labrum
repair in patients 40 years old and greater. This is a clinically important question for both physical
therapists and surgeons because of their ethical duty to provide patients with the most appropriate
care. It is also relevant to health care providers because it can help eliminate certain waste within
their profession. Specific inclusion criteria include patients 40 years or older, magnetic resonance
imaging (MRI) evidence of acetabular labral tear, history of consistent symptomatic labral tear, a
flexion, adduction, and internal rotation (FADIR) or flexion, abduction, and external rotation
(FABER) test to confirm diagnosis, and willingness to be randomized. The study was conducted
using one surgeon, a group of independent physical therapists, and a host of outcome measures used
to assess functional abilities of a hip joint. A modified as-treated analysis and a treatment-failure
analysis was recorded in this study. Results of this study indicate that it is possible that surgical
treatment for these types of acetabular labral tears with subsequent physical therapy is more
effective than physical therapy programs alone. This study can be used to assist physical therapists
to refer to surgeons on a more appropriate basis, surgeons to follow surgery with physical therapy,
and for the overall increased quality of life within patients.

1. Was true randomization used for assignment of participants to treatment groups?

a. The study used an electronic randomization program to allow an allocator to

randomize participants to either surgery plus physical therapy or physical therapy
alone in a 1:1 ratio. This article used a simple random sample method because each
participant had an equal chance of being placed in either group. Furthermore, the
distribution of the subject characteristics was equal across both groups.

2. Was allocation to groups concealed?

a. The article mentioned that an allocator randomized the groupings, but it did not
explicitly state if they were aware of the treatment/control group before. This can be
dangerous to the study because of the risk of the allocator purposely assigning people
to a group that they think will fit the study most. This could be an instance of
confirmation bias because the allocator could have been trying to sway the results to
align to their opinion. This would distort the results and destroy the efficacy of the
study.

3. Were treatment groups similar at the baseline?

a. It was unclear if they were similar or not. As mentioned in the first question, the groups
were all similar in terms of a clinical significance of their hip labral tear. All the
participants had either a FADIR or FABER test and an MRI to confirm a diagnosis of
a hip labral tear. In these terms, they were similar; however, there is no mention of
other factors that may make them similar. Factors like sociodemographic, prognostic
factors of the tears, and addressing any potential differences were not mentioned. The
article could have had a better description of the similarities and differences between
the groups.
© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.
4. Were participants masked (or blinded) to treatment assignment?

a. Due to the study’s design, no blinding occurred. The study addressed this and
mentioned that they would have liked to use a blind treatment assignment, but it was
impossible because of issues of consent and timing to have surgery.

5. Were those delivering treatment blind to treatment assignment?

a. For the same reasons as the previous question, the article mentioned that the single
surgeon performing the surgery could not have been blinded. In a similar manner, the
physical therapists delivering care could also not have been blinded either because
they would have to follow different protocol for post-operative care.

6. Were outcomes assessors blind to treatment assignment?

a. Blinding was not possible in this study due to the comparison of surgical and
nonsurgical interventions. Although they were not blind, the outcome assessors were
the individual physical therapists that delivered treatment, making them a qualified
source.

7. Were treatment groups treated identically other than the intervention of interest?

a. The groups seem to have been treated identically other than the manipulation of the
study. The groups were obviously different in the fact that one had surgery and the
other did not, but they also had different protocols. The two groups had to have
different protocols to protect the surgical repairs, thus this does not qualify as treating
the groups differently. Although it is necessary to use different protocols, it is possible
the differences in protocols could have caused better outcomes for one group. Finally,
there was no evidence of any other differences in treatment style between the groups.

8. Was follow up complete and if not, were differences between groups in terms of their follow
up adequately described and analyzed?

a. Yes, the article included a follow up. The study included a 90% follow-up rate after
enrolling 90 participants, and 81 of them completed the 12 month follow up protocol.
The study included the reasons for dropouts, including 7 who did not follow their
protocol and 2 who decided to simply withdraw from the study. The article also
included that 66% of the participants who were in the physical therapy alone study
crossed over to the surgical option after a physical therapist determined there was no
more functional gain from therapy alone. Overall, the study did a good job of including
all the data, good or bad, in the results.

9. Were participants analyzed in the groups to which they were randomized?

a. An intention-to-treat analysis was performed and mentioned within this study. The
article reports that due to a high cross over rate, multiple sensitivity analyses were
used. This included a modified as-treated approach with the expectation of using a
time-varying fixed effect for the treatment groups. A treatment-failure analysis was

© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.
 also used as a sensitivity analysis. A detailed CONSORT flow diagram was provided
within the article.

10. Were outcomes measured in the same way for treatment groups?

a. The outcomes were measured the same way between the two groups. The study used
the international hip outcome Tool (iHOT-33) and the modified harris hip score
(mHHS) as the major assessors. Secondary outcome measures such as hip outcome
score activity of daily living (HOS-ADL), sport subscale (HOS-SSS), nonarthritic hip
score, lower extremity functional scale, and visual analog pain scales were used as
well. These outcome measures were given before the start of treatment, 3, 6, and 12
months; however, the iHOT-33 and mHHS were only used at 12 months after
randomization.

11. Were the instruments used to measure outcomes reliable and valid?

a. Raters were the individual therapist assigned to each patient using an intra-rater
approach. It mentions in the article that the team originally had a different outcome
measure test but switched to the iHOT-33 because it was deemed more reliable and
valid in other studies. The fact that the group of experimenters thought to use a
different outcome measure based on new data shows that they are aware of the
continuously evolving medical profession and want to provide the most accurate data.

12. Was appropriate statistical analysis used?

a. The dependent variables within the study were all the outcome score measures. The
independent variable of the study was surgery and age. This was an appropriate choice
because the study was intended to view the results of surgical intervention and physical
therapy vs physical therapy alone for diagnosed femoral acetabular labrum tears in
patients 40 years and older. The effect size was calculated within the article as well.
Finally, the statistical analysis was explained well throughout the article, giving
reviewers plenty of information on why certain analyses were chosen and the results
from them.

13. Was the trial design appropriate for the topic, and any deviations from the standard RCT
design accounted for in the conduct and analysis?

a. As previously mentioned, deviations had to be made due to the nature of the study.
Such deviations included having to use a cross over design due to some patients being
deemed to have no further medical progressions with physical therapy alone. The
results of this study were determined to be statistically meaningful, and they would be
useful in the clinic for multiple reasons. If a patient were to present with signs and
symptoms consistent with those of this study, it may be more beneficial in terms of
time, money, and quality of life to refer a patient to a surgeon before treatment is given.
Being efficient with these factors are extremely important in physical therapy, which
makes this study worth the cost of investment.

© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.
 Additional consideration

Previous studies of the same nature tend to show the same results as the study I reviewed;
however, there is not a lot of information about patients over 40 years old, making this article
even more important. This journal and authors are well-known in the sports medicine community,
making it a reliable source of information. The authors have each been published in multiple other
articles, most of them relating to studies of a similar nature.

Why you should or should not use this evidence?

Although certain factors were not included into this study, I still think this should be used as a
strong source of evidence. My biggest concern would be that the article did not make it extremely
clear of the similarities/differences between groups at baseline. This could potentially force
reviewers to think results could be skewed because certain factors other than the independent
variables could have contributed to the results. Any concerns of blinding can be dismissed because
there are ethical and legal concerns with blinding participants when surgery is involved. Overall,
I think this article could have been improved; nonetheless, I would still consider it to have
moderately strong evidence.

© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.
© JBI, 2020. All rights reserved. JBI grants use of these tools for research purposes only. All other enquiries
should be sent to jbisynthesis@adelaide.edu.au.