JBJS welcomes reader comments on published articles. Letters to the Editor are reviewed by JBJS editors but are not peer-reviewed. To submit your letter, please follow the "submit a response" link that appears in the content box at the upper right of the full text of the article.

Letters to the Editor to:

Scientific Articles:
Arabella I. Leet, Carmen P. Pichard, and Michael C. Ain
Surgical Treatment of Femoral Fractures in Obese Children: Does Excessive Body Weight Increase the Rate of Complications?
J Bone Joint Surg Am 2005; 87: 2609-2613 [Abstract] [Full text] [PDF]
*Letters to the Editor: Submit a response to this article

Electronic letters published:

[Read Letter to the Editor] Dr. Leet , et al respond to Dr. Lubicky
Arabella I. Leet, M.D., Carmen P. Pichard, M.D., Michael C. Ain, M.D.   (22 February 2006)
[Read Letter to the Editor] Clinically Insignificant Results
John P. Lubicky, M.D., FAAOS, FAAP   (17 January 2006)

Dr. Leet , et al respond to Dr. Lubicky 22 February 2006
Previous Letter to the Editor Top
Arabella I. Leet, M.D.
Johns Hopkins University, Baltimore, MD,
Carmen P. Pichard, M.D., Michael C. Ain, M.D.

Send letter to journal:
Re: Dr. Leet , et al respond to Dr. Lubicky

aleet1{at}jhmi.edu Arabella I. Leet, M.D., et al.

We appreciate Dr. Lubicky’s comments and would like to address his concerns. Dr. Lubicky feels that in our paper we are addressing our “anecdotal feelings” about operating on obese children, yet there are reports in the current orthopedic literature that support the idea that complications are greater and operative times are longer for obese adults when compared to non-obese adults. Jupiter, et al,(1) reported complications related to positioning of heavy limbs during surgery for fracture non-unions in obese patients. Problems with positioning, combined with an increase in operative time, resulted in nerve palsies, compartment syndromes, and scalp alopecia (1).

A second study examining the treatment of femur fractures in obese adults demonstrated a high rate of wound dehiscence and thromboembolic disorders complicating surgery (2). Thus our aim was to determine whether the same complications and increased operative times seen in obese adults were occurring in the pediatric population.

The paper does have a relatively small number of obese patients (the discussion section explores some of the reasons for this). Our initial analysis had contained a larger cluster of children in the obese group with a cut off at the 90th percentile of weight for age. Upon review of our manuscript, the Journal of Bone and Joint Surgery asked us to use the Center for Disease Controls’ more rigorous definition of obesity at the 95th percentile of weight for age. We were concerned that there would be a loss of statistical significance when the group we defined as obese became smaller. Instead, we found that the number of complications was still large enough to result in statistical significance, even with the more rigorous definition of obesity. Thus the opposite of Dr.Lubicky’s concern occurred—-when we removed many patients from the obese group, we still maintained statistical significance.

Our definition of complications was not a draining pin, or a course of oral antibiotics for cellulitis, but instead complications that resulted in readmission to the hospital or reoperation. These types of complications resulted in lost time from school for children and work for parents. We could have easily increased the number of complications by including some minor nuisances, but we did not feel that this was appropriate.

We agree that using estimated weight is a weakness of the paper. The study was retrospective and most of the patients arrived in our emergency room in hare traction—unwilling and unable to move. The weight estimate placed in the chart was a caretaker’s estimated weight. This weight became the “working” weight, and many decisions such as amount of anesthetic given, or dosages of pain medications, or antibiotic dosages were based on this weight. We know of no patient who was not medicated properly based on the estimated weight. In some cases we had precise weights; we just did not have this weight information on everyone.

While statistical significance is easily defined as a p value of less than 0.05, clinical significance is a more subjective concept. We found a 12% complication rate in normal children, a 25% complication rate in heavy children, and a 50% complication rate in obese children (p<.004). The ability to detect a difference between these groups using the strictest definitions of both obesity and complications leads us to feel confident that our results (a 4-fold increase in complications in obese children)are both clinically and statistically significant.

References:

1. Jupiter J, Ring D, Rosen H: The Complications and Difficulties of Management of non-unions in the severely obese. J Orthop Trauma 9 (5): 363 -70, 1995.

2. McKee MD and Waddell JP: Intramedullary Nailing of Femoral Fractures in Morbidly Obese Patients. J Trauma 36 (2): 208-10, 1994.
Clinically Insignificant Results 17 January 2006
Next Letter to the Editor Top
John P. Lubicky, M.D., FAAOS, FAAP,
Othopaedic Surgeon
Indiana University, Medical Ctr/Riley Hospital for Children, Indianapolis, IN

Send letter to journal:
Re: Clinically Insignificant Results

jlubicky{at}iupui.edu John P. Lubicky, M.D., FAAOS, FAAP

To The Editor:

This letter is in response to an article by Leet et al (1)that reports complications in obese children undergoing surgical treatment of femoral fractures. While most of us surgeons, it is safe to say, prefer to operate on thin patients rather than those who are obese, most of us have the anecdotal impression that various kinds of complications seem to be more common in obese patients whether they are children or adults. The authors apparently wanted to prove that these anecdotal feelings are actually true and could be quantified statistically by looking at a group of children undergoing surgery for femoral fractures. Unfortunately, I don't believe that they proved their point for a variety of reasons.

The first problem is that there are a relatively small number of what they have classified as obese patients among the total number of patients in the study, and for this reason it is likely that adding or subtracting one patient from that small group, particularly one that has a complication, would really change the statistics quite significantly. At least it would change the percentage of the prevalence of the complications within this group. Statistical analysis of the findings would seem to lack significant power because the number of patients is so small. When faced with studies like this with small numbers of patients, at least for me, the statistical significance is always somewhat questionable.

The second issue that I have with the paper is that the key element, that is the patient's weight, was inaccurately determined according to the article and it was mainly an estimate, for the most part, made by the anesthesia staff and/or a parent. One did not get the impression that the child was actually weighed on admission to the surgical unit preoperatively. In addition, no height measurements were made and therefore, the authors were limited in terms of how they could determine the child's size relative to some standard. They had to use the "weight per age" classification rather than the BMI (Body Mass Index).

Therefore, the authors were limited in their ability to classify these patients into groups based on their size. Although the BMI is not totally without its problems, it is probably the "gold standard" that is used in various disciplines to reflect the size of a person and it is certainly used commonly to denote degrees of obesity.

Only 10% of the total study group was considered "heavy." The authors further divided the "heavy" patients into those who were "obese" using the Centers for Disease Control and Prevention criteria, which require that a child be above the 95th percentile in weight per age. A second subgroup of "heavy", comprised of children who were "very heavy" and who were between the 90th and 94th percentiles. Again, by doing this, two even smaller groups of patients were created. With such small numbers, one questions the utility and significance of the data.

The types of complications that occurred in the heavy patients were generally associated with wound infection or a wound dehiscence. Only one of the heavy patients had what could be considered a much more significant complication which was refracture through an old fracture site. None of the heavy patients had other substantial complications such as loss of alignment, malunion, osteomyelitis, compartmental syndrome, or broken rods or pins. In fact, the combined heavy and non-obese patient groups had the same incidence of pin tract infections as the non-obese and the non-obese had twice as many refractures. The data also indicated that the operative time was longer for the non-obese patients.

The publication of this paper contradicts the very rhetoric included in discussions on levels of evidence (2-5). The very essence and critical data that underly the conclusions reached by the authors of this paper rely on an accurate measurement of the patient's weight and the authors freely admit that the weight they used for their calculations was only an estimate. Additionally, the number of patients was so small as to render conclusions that were statistically significant, clinically insignificant.

In the end, based on data present in this article, it seems as though our anecdotal experience with complications in obese patients may actually have been proven wrong since none of the obese patients had the more serious kinds of complications following surgery for femur fractures that in fact might very well require reoperation. So much for levels of evidence and evidence-based medicine.

References:

1. Leet AI, Pichard CP, Ain MC. Surgical treatment of femoral fractures in obese children: does excessive body weight increase rate of complications? J Bone Joint Surg Am. 2005;87:2609-2613.

2. Carr AJ. Evidence-based orthopaedic surgery, what type of research will best improve clinical practice? J Bone Joint Surg (Br.) 2005;87- B:1593-1594.

3. Horan FT. Judging the evidence. J Bone Joint Surg (Br.) 2005;87- B:1589-1590.

4. Obremskey WT, Pappas N, Attallah-WasifE, Tometta P 3rd, Bhandari M. Level of evidence in orthopaedic journals. J Bone Joint Surg Am. 2005;87:2632-2638

5. Tovey D and Bognolo G. Levels of evidence in the orthopaedic surgeon. J Bone Joint Surg. (Br) 2005187-B:1591-1592