Is innovation in surgery less than ideal? A case study of acellular dermal matrix assisted prosthetic breast reconstruction

Introduction

The introduction of innovative procedures requires appropriate evaluation. IDEAL recommendations propose four stages of evaluation, (I dea-D evelopment-E valuation-A ssessment-L ong-term study). The aim of this study was to review the introduction of an innovative surgical technique according to this framework.

Methods

Literature searches identified articles published between 2000 and 2012 reporting acellular dermal matrix-assisted prosthetic breast reconstruction (ADMPBR). Studies were classified by IDEAL-stage as A)descriptive (IDEAL-1/2a) reporting the feasibility or development of ADMPBR or B)comparative (IDEAL-2b/3(RCTs) comparing ADMPBR with standard techniques. IDEAL study designs reported before and after 2008/9 were examined to explore progression of study design over time.

Results

Of 236 abstracts, 50 papers reporting data on 3,648 patients were included. 24 (48.0%) were IDEAL-1/2a;25 (50.0%) IDEAL-2b and 1 (2%) IDEAL-3. IDEAL-2b-studies significantly increased from period-1 (2005-2008) to period-2 (2009-2012)(n=1 to n=24,p<0.01). The number of IDEAL-1/2a-studies published annually remained constant (n=2-4). Almost all IDEAL-1/2a studies (n=20,87.0%) provided comprehensive descriptions of surgical technique, but less than half (n=11) reported patient-selection criteria and only 25% documented seeking patient consent IDEAL-2b-studies were significantly larger than IDEAL-1/2a-studies (IDEAL-1/2a-median= 39, inter-quartile range-20-65 vs.IDEAL-2b-median=73, inter-quartile range-36-186,p<0.01,Median-test) and more likely to report combined results from groups of surgeons (n=10 vs.n=5;p=0.06). Short-term complication reporting was more comprehensive in IDEAL-2b-studies but there were no differences in the reporting of histological or technical details across groups and IDEAL-1/2a-studies were significantly more likely to report long-term (p=0.03), patient-reported (p<0.01) and cosmetic outcomes (p=0.05).

Conclusions

The introduction of ADMPBR does not consider previous evidence and comparative studies are lacking. Well-designed and conducted studies are needed to appropriately evaluate novel surgical innovations to establish standards of care, protect patients and surgeons.