Feb 2011
Plastic & Reconstructive Surgery
Commentary
19 July 2000
Robert Goldwyn, MD Editor, Plastic and Reconstructive Surgery 1101 Beacon Street Brookline, MA 02146
Dear Sir,
Thank you for the thoughtful and interesting remarks made by Carolyn L. Kerrigan and E. Dale Collins in their letter to the editor, "Are Perforator Flaps Truly More Cost-effective Than TRAM Flaps? How Good is the Evidence?" The authors clearly put much time into their lengthy and well-cited letter, but we're afraid that we cannot agree with their final conclusion.
While the authors made important observations in their letter, they appear to have missed the point of our article. Our study was an observational comparison of two institutions with comparable volumes and surgical experience in two different procedures, the free perforator flap and the free TRAM flap. In fact, with regard to experience and volume and its effects on efficiency, at the time of our study we at LSU had less than five years experience with the free perforator flap while those at MD Anderson had well over five years experience with the free TRAM flap. If anything, we were at a disadvantage. Since that time, the surgeons at MD Anderson have, in the interests of optimal patient care, replaced the free TRAM flap with the free perforator flap. Kerrigan and Collins themselves admit that while they do dispute our paper, they "do not dispute that perforator flaps may be more cost effective."
In fact, one reason for the cost-effectiveness is the shorter stay of patients who have had a perforator flap breast reconstruction. As stated in our article, the reason for this is that underlying muscle tissue is left intact in the perforator flap reconstruction, and the inevitable result is a drastically lowered morbidity rate in essentially undisturbed tissue. Given this, we would not be able to support the authors' contention that our study demonstrates that a shorter stay might be possible for all free flap breast reconstructions, including those in which muscle tissue was sacrificed. Instead, we would argue that like implants before it, the myocutaneous flap breast reconstruction may be outdated, and should be considered a second-line treatment for breast reconstruction. Indeed, patients should be informed of their right to preserve their muscle tissue in autologous breast reconstruction, and the myocutaneous flap should only be used when it is necessary to transfer muscle tissue to restore function. In all other cases, the free perforator flap breast reconstruction must be considered the new standard of care.
Robert J. Allen, MD Constance Chen 4429 Clara Street Suite 440 New Orleans, LA 70115
.............................................................................................................................
June 02, 2000
Robert Goldwyn, M.D. Editor, Plastic and Reconstructive Surgery 1101 Beacon Street Brookline, Mass 02146
Dear Sir,
We cannot agree with the conclusions made by Jonathan Kaplan and Robert Allen in their paper, "Cost-Based Comparison between Perforator Flaps and TRAM flaps for Breast Reconstruction " (Plast. Reconstr. Surg. 105 (3): 943, 2000).(1) Although there are important conclusions that can be drawn from this manuscript, they are not the ones suggested by the authors. Principally, there are serious flaws in their concluding statement that "this technique should gain credibility and greater use in breast reconstruction" based on their purported findings of shorter hospital stay, shorter surgery and thus lower cost.
Others have developed broad guidelines for rating medical literature based on study design, allowing readers to weich the quality of the evidence. Level 1 studies provide the strongest evidence and level 5 the weakest (Figure 1 (2)). As a case series, this paper by Kaplan and Allen would be graded as a level IV study. Further limitations are its retrospective nature and its lack of a comparison group built into the study design. While the authors look to the literature for an historical control, (3) there are inherent flaws in this approach, as many variables cannot be controlled for. At a minimum, these uncontrolled variables, which may in themselves account for any differences seen, should be acknowledged. Therein lies the weakness in this paper. Although Kaplan and Allen address and adjust for change in value of the dollar between 1993 and 1997, there are multiple other potential variables that they fail to enumerate and address. In particular, if the observed differences in cost are real, then what are the alternative explanations for these differences? While we regard duration of surgery and length of hospital stay as reasonable proxies for cost, there are other factors that must be considered. Some obvious elements that could account for the observed differences include:
In their study, Kaplan and Allen based their analysis on 59 DIEP flaps and 5 gluteal artery perforator flaps. The steps performed in carrying out a free flap breast reconstruction can be broken into 3 steps: 1. harvesting of the free flap, 2.completion of microsurgical anastomoses and 3. sculpting of the breast. The latter two are essentially identical between the TRAM and the DIEP flap. Therefore, it is reasonable to assume that it is the harvesting of the flap that accounts for differences in time to complete either approach. The harvest of these two flaps is quite similar with the exception of the final step in muscle dissection versus perforator dissection. In experienced hands, a surgeon is faster at the technique with which they are most familiar. In learning a new technique, one goes through a leaning curve to develop equal speed and efficiency. Having said this, if Kaplan and Allen had data to compare their own speed at doing TRAM's versus DIEP'S, we would be surprised if there was a difference in surgical time of 2 hours.
Other surgeons performing perforator flaps have reported mean operating 4 times of 6.2 hours (4) and 4.67 hours (5) for unilateral reconstruction and mean hospital stay of 7.9 days.(4) The experience of both of these authors led to comments such as "the more complex nature of this type of surgery, leading to increased operating time" and "the more tedious flap dissection" These reports emphasize the known variability in operative times and fallacy inherent in Kaplan and Allen's conclusion. A more recent abstract from the senior author's group indicates that their current surgical time has decreased further to 5.4 hours while their hospital stay has increased to 4.7 days! (6) Operative times vary by procedure, however the procedure is but one of many factors responsible for variations.
Surgical volume is directly correlated with efficiency, speed, lower complications and better outcomes. This has been demonstrated in numerous large surgical series, where the relationship between surgical volume and outcome such as mortality and complications has been demonstrated convincingly.(7,8) This has also been demonstrated in our literature where Kroll et. al. showed shorter operative times and better outcomes later in their series of free TRAM flaps.(9) It follows that increasing volume of a given procedure is likely to have an impact on efficient use of OR time and length of hospital stay. If Kaplan and Allen chose to compare their experience with that of another surgeon performing perforator flaps, they would likely have found differences in duration of surgery and length of stay. Likewise, if they had chosen high volume surgeons doing TRAM'S, they may have found the TRAM to take less surgical time and have a shorter hospital stay than their current practice with perforator flaps. Anecdotally, we are aware of teams performing free TRAM breast reconstructions in 4 hours or less, with hospital stays of 3 days.
We are all aware that more than surgical technique drives operative time and length of hospital stay. For example, some general surgeons prefer to complete the mastectomy before allowing the plastic surgeon to begin the reconstruction, whereas others encourage flap elevation simultaneously with the mastectomy It is not clear from the papers being compared if the sequencing of surgical care is similar. If not, this alone could account for the difference in duration of surgery rather than the technique. The size and skill of the surgical team is also a critical determinant. Most surgeons rely on trainees or nonphysician surgical assistants in completing, TRAM'S. We don't know what the full composition of the team was in this paper, however, there are at least two plastic surgeons working together on the DIEP flaps. Certainly this would appear to be a more expensive, but more efficient approach, than the norm of practice. Yet Kaplan and Allen fail to address this issue in their analysis. Even if the direct salary costs were accounted for, the lost opportunity costs would be significant.
Though we cannot agree with the conclusions drawn by Kaplan and Allen, their work does contribute useful information to the literature. Firstly, they have established a standard suggesting that it may be safe to discharge a free flap after only 3 days of hospitalization. This is in contrast to a style of practice characterized through the '80's and early '90's by patients being routinely hospitalized for 7 days (3) and more recently shortened to 5 days (10). There are even suggestions that outpatient free flaps may be an option for some indications.(11) Hwang et. al. implemented a clinical pathway and documented an ability to shorten hospital stay without incurring an increase rate of complications.(10) If Kaplan and Allen were to share the details of the postoperative management of their patients in the form of a clinical pathway, many of us would likely benefit. In any case, this report represents a valuable benchmark for others to work towards.
Secondly, Kaplan and Allen also have set a published benchmark of 6.2 hours for breast reconstruction with a free flap. However, it is not clear if the duration reflects their experience and high volumes, the use of a perforator flap rather than TRAM, the expertise of the surgeons or support staff, or the sequencing of simultaneous flap elevation during the mastectomy. As a group, physicians have a tendency to assume ultimate responsibility for clinical outcomes. This unjustly minimizes the importance of the environments in which we work and the role played by the total care process. Other team members (anesthesia, hospital nursing staff, community nursing staff, and family members) and their attitudes also have a profound influence on the efficiency with which we work and the outcome of our interventions.
We do not dispute that perforator flaps may be more cost effective, but we do dispute that this paper produces the evidence to support that conclusion. The conclusion that these data support are that indicators of quality care (such as duration of surgery and length of stay) for similar procedures can vary widely from institution to institution. The real service that these authors provided was in demonstratincy that free flap breast reconstruction can be done with shorter stay and this should encourage us all to examine our practices and strive for efficiencies and outcomes that are comparable to the leaders in our field. The authors are challenged to come up with more rigorous scientific data to support their conclusion. In closing, we would like to encourage readers of the journal to take a careful look at the design of studies in terms of their quality and weight of evidence before incorporating new recommendations into their own practice.
Carolyn L. Kerrigan, M.D. Professor of Surgery One Medical Center Drive Lebanon, NH, 03756 [email protected]
E. Dale Collins, M.D. Assistant Professor of Surgery
Table 1:LEVELS OF EVIDENCE (Adapted from the American Society of Clinical Oncology) 2
Level 1
Randomized controlled trial or meta analysis
Level 2
Excellent experimental study / small randomized controlled trial
Level 3
Quasi experimental study / case-control series
Level 4
Descriptive and case studies
Level 5
Case reports / clinical examples
References