Opsomer G, Deleu PA, Bevernage BD, Leemrijse T. Cortical thickness of the second metatarsal after correction of hallux valgus. Foot Ankle Int.  2010 Sep; 31(9): 770-6. (PubMed ID: 20880479)

WHY did the authors undertake this study?
A long held supposition within foot and ankle surgery is that the hallux valgus deformity and first ray instability can lead to a transfer of stress to the second metatarsal as manifested by cortical thickening.  The simple, and unique, aim of this study was to evaluate if surgical correction of hallux valgus had any effect on the cortical thickness of the second metatarsal.

HOW did they attempt to answer this question?
The primary outcome measures of the study were radiographic parameters of medial 2nd metatarsal cortical thickness, lateral 2nd metatarsal cortical thickness, 2nd metatarsal intramedullary thickness, entire shaft thickness of the 2nd metatarsal, entire shaft thickness of the 4th metatarsal (used as a control) and the “MCT/ST” ratio which the authors defined as the mean medial cortical thickness divided by the entire shaft thickness of the metatarsal. 

Standardized AP radiographs were evaluated in 13 feet before and after hallux valgus corrective surgery (mean follow-up 16 months; range 11-24 months). . 

WHAT were the specific results?
Statistically significant differences were found  post-operatively with respect the the medial 2nd metatarsal cortical thickness (decreased 0.16mm) and 2nd metatarsal intramedullary thickness (increased 0.27mm).  No significant differences were found with respect to the entire thickness of the 2nd metatarsal shaft, the lateral 2nd metatarsal cortical thickness, or any measure involving the 4th metatarsal.

HOW did the authors interpret these results?
From these results, the authors concluded  that surgical correction of the HAV deformity may lead to a redistribution of stress across the 2nd metatarsal.

Belch JJ, Dormandy J; CASPAR Writing Committee. Results of the randomized, placebo-controlled clopidogrel and acetylsalicyclic acid in bypass surgery for peripheral arterial disease (CASPAR) trial. J Vasc Surg.  2010 Oct; 52(4): 825-33. (Pubmed ID#:  20678878)

WHY did the authors undertake this study?
Recent studies have demonstrated the beneficial effects of a combination of acetylsalicyclic acid (ASA) and clopidogrel anti-platelet therapy in patients with a cardiovascular event or receiving percutaneous coronary interventions as compared to ASA alone.  From these results, the authors hypothesized that this combination anti-platelet therapy would be more beneficial than ASA alone in patients undergoing below-knee bypass grafting for the treatment of peripheral arterial occlusive disease.

HOW did they attempt to answer this question?
The primary outcome measures of this study  were patient endpoints including graft occlusion, need for ipsilateral leg revascularization, ipsilateral leg above-ankle amputation and death.

A prospective, multicenter, double-blind, placebo-controlled study was utilized to initially randomize 851 patients to receive either clopidogrel plus ASA or ASA alone following unilateral below-knee arterial bypass grafting.

WHAT were the specific results?
There was no statistically significant difference between the groups with respect to any of the primary patient endpoints.  With subgroup analysis however, reaching a primary endpoint was significantly reduced with dual anti-platelet therapy when prosthetic grafts were utilized for the bypass (but not with venous grafts).  No significant difference in bleeding risk was appreciated between the two groups.

HOW did the authors interpret these results?
From these results the authors concluded that post-operative dual antiplatelet therapy did not improve patient outcomes in patients requiring below-knee bypass grafting, but may have some benefit in patients receiving prosthetic grafts.

Let’s take a closer look at the topic of measurement precision, particularly as it applies to the Opsomer et al article.  The authors of this study asked a very interesting and unique clinical question with respect the pathophysiology and surgical correction of the HAV deformity.  One of their primary results and conclusions was that the thickness of the medial cortex of the 2nd metatarsal significantly decreased following HAV surgical correction.  They report that the average (standard deviation) preoperative medial cortical thickness was 2.66mm (0.40), and postoperatively this decreased to 2.50mm (0.40).  Using a student paired t-test to analyze and compare these values (an appropriate test assuming the data was drawn from a normally distributed population), they found that this was a statistically significant decrease with a p < 0.001. 

Whenever a statistically significant difference is reported, it is always useful as a critical reader to take the analysis a step further and asked if the difference is also clinically significant.  In this case the actual, real number decrease in cortical thickness was 0.16mm.   Take a minute and think about how small a millimeter is…and then think about attempting to divide that millimeter into 1/10s!   This is the actual, clinical difference that we are dealing with here.   This decrease represents only about 2% (0.16/7.8) of the overall thickness of the 2nd metatarsal in this study, and (interestingly) falls well within the standard deviation of each of the pre- and post-operative measurements (0.40mm). 

The point I want to bring up however is whether or not this difference represents an actual measurable difference, or to put it another way, whether or not this difference falls within the measurement precision.  Anytime we take any type of measurement, we need some type of device to actually take the measurement.  And any measurement device has an inherent amount of precision to it. Take another minute and grab a ruler or a goniometer and take a close look at the gradations.  Most rulers are precise enough to take a measurement to the nearest millimeter, just as most goniometers are precise enough to take a measurement to the nearest degree.  But would you feel comfortable using this ruler to take a measurement to the nearest 1/10 of a millimeter, or to take it even a step further and measure to the nearest 1/100 of a millimeter (like 0.16)?  I certainly wouldn’t, and I think most people would agree with me!

Now the Opsomer et al authors did not use a standard ruler to take their measurements of course.  In fact, they used a computed/digital radiograph system.  But even this type of system has an inherent precision to it, and all measurements are dependent on the two points where the authors “click” their mouse to define the measurement.  The authors did a good job to define for use the intra- and interobserver reliability of their mouse “clicks”, but we are provided with no information with respect to how “precise” their measurement system was.   It may very well be that their precision was to the nearest 1/10 or 1/100 of a millimeter, but this reservation, combined with the limited clinical significance of the findings, makes me a little less enthusiastic about the conclusions of this study as a critical reader.