The patients scheduled to undergo a primary ACL reconstruction surgery at a university medical center in Korea were enrolled consecutively in this study from January 2007 to December 2008. Patients who had undergone prior knee ligament surgery on either knee, or had multiple ligament injuries were excluded. One hundred and twenty-nine patients were enrolled this study. Written informed consent was obtained from all patients. All patients had a preoperatively documented ACL rupture or a deficiency detected by magnetic resonance imaging examinations. The preferred technique for ACL reconstruction in our clinic is an arthroscopic procedure using either quadruple hamstring tendon autografts or tibialis anterior tendon allografts. An independent investigator not associated with the study counseled the patients regarding the surgical procedures and the risks and benefits of each graft type during outpatient interviews. He attempted to be as neutral as possible regarding the graft choice and spent an equal amount of verbal explanation time on each graft type. Patients also received educational leaflets containing general ACL graft information with descriptions of hamstring tendon autografts and tibialis anterior tendon allografts (Fig. 1). The patients ultimately decided which graft type would be used. The participants included 113 men and 16 women with a mean age of 34.3 years (range, 15 to 57 years). The participants completed an 11-item questionnaire discussing their graft decisions on hospital admission for surgery (Fig. 2).

An ACL reconstruction was performed on each participant using the patient-selected graft. All arthroscopic procedures were performed by a single surgeon. The surgical procedures in autograft and allograft groups were identical except for the nature of the graft. Endobutton (Smith & Nephew, Memphis, TN, USA) was used for femoral fixation and a double fixation technique (the spike washer/screw and the bio-absorbable interference screw) was employed for tibial fixation in both groups.

The data was analyzed using SPSS ver. 13.0 (SPSS Inc., Chicago, IL, USA). The continuous variables are reported as the mean standard deviation and the categorical variables are described using frequency distributions and reported as percentages. The normally distributed continuous variables were analyzed using a Student's t-test. The categorical variables were examined using a chi-square or Fisher's exact test if the values in the cells were 5. A p-value < 0.05 was considered significant.

The bone-patellar tendon-bone (BPTB) and hamstring tendon autografts are the most common grafts in ACL reconstructions.1) The BPTB autograft is frequently chosen because of its excellent clinical results and high level of patient satisfaction after a long term follow-up.4,9) Some studies suggested that a harvest of the central third of the patellar tendon might have associated donor site morbidity, such as patellofemoral osteoarthritis, patellar tendon shortening, loss of terminal extension, and patellofemoral pain.5,10,11) There has been an increase in the popularity of hamstring tendons as autografts for ACL reconstructions, which can avoid harvest site morbidity.1) However, Aune et al.12) reported that hamstring tendons exhibited inferior flexion strength at the 24 month follow-up. Debates over graft selection are likely to persist with the presence of autograft harvest site morbidity.

The potential advantages of allograft placement include autograft benefits without donor site morbidity. Krych et al.7) performed a meta-analysis of autograft and allograft stability data, and reported that the BPTB autograft procedure had less graft failure and a better functional outcome (as measured by the single leg hop test). However, all measurable outcomes were similar when irradiated and chemically processed grafts were excluded from the analysis. This suggests that the allograft choice is dependent on cost, the potential for disease transmission, sterilization protocols and their impact on graft integrity and strength, immunogenicity and clinical results.

Surgeons typically choose grafts for ACL reconstructions according to their experience, preference and tissue availability. It is believed that many therapeutic decisions need to be individualized, particularly when they involve choices between possible outcomes that may be viewed differently by different patients. The patient preferences are an essential element of care giving,13) but there has been minimal research on the patient preferences on graft selection in the literature. To our knowledge, no previous study has allowed patients to choose their own graft types and examined the factors affecting their decisions. In the present study, the patients chose their own grafts to allow greater patient participation in surgical care.

ACL reconstructions are covered by the National Health Insurance in Korea. However, allograft use is covered by insurance only if autograft sources are unavailable. The patients subsidize extra allograft charges if this procedure is selected electively. Autograft ACL reconstruction costs approximately US$1,500, whereas an allograft ACL reconstruction costs approximately US$3,000. The ACL reconstruction patients were hospitalized for 3 days postoperatively regardless of the graft type. The majority of participants in this study selected autografts. None of the participants asked the surgeons to choose their graft type or were hesitant to make their choices. Although the socioeconomic status of the autograft and allograft groups could not be compared directly, the data suggests that the cost of an allograft in Korea is a significant deterrent to its selection. However, there were no significant differences in education levels between the two groups. The results may vary in regions or countries where the costs are comparable.

The surgeon's explanation (with educational leaflets) was the most important source of information contributing to the choice of graft in both autograft and allograft groups. However, there was a significant difference in the proportion of sources of information between the groups. The use of Internet-derived health information has increased rapidly and has been termed the "e-patient revolution."14) Gerber and Eiser15) reported that Internet-derived health information may provide additional information to patients, lead to better health outcomes, and allow a more appropriate use of health service resources. On the other hand, the quality of information on the Internet can be highly variable and may limit its use as a visible information source.16) More patients in the allograft group stated that the Internet affected their graft decisions. Internet searches using the keywords "anterior cruciate ligament," were conducted using the three most popular search engines in Korea (www.naver.com, www.daum.net, kr.yahoo.com). The sponsored links, blogs, and news were reviewed. Information on the ACL anatomy, injury, surgery, graft type, and rehabilitation was acquired, but were unable to locate sites showing the superiority of one graft type over another.

This study had several limitations. First, the educational leaflet for graft information and the questionnaire evaluating the patients' choice were developed in house. There is some concern as to whether the leaflet and questionnaire introduced bias. These documents may require revision by professional societies in the future. Second, regardless of how carefully the investigator tried to be neutral regarding the graft choice and spend an equal amount of verbal explanation time on each graft type, his own personal bias might have been reflected in his presentation, which may have guided the patient toward one of the two choices. Third, the patients were offered only one of two different graft choices. There are many other graft choices reported in the literature, including quadriceps tendon, BPTB, Achilles tendon allograft, and other possible options. The patient's decision was limited to the graft type and no choices regarding other surgical options were offered. It is believed that autogenous tissue harvest and allogeneic tissue use are critical issues to specific individuals, and decisions regarding other surgical portions of ACL reconstruction should remain the surgeon's choice.

In conclusion, patient graft selection is a reasonable way of designating the type of surgical procedure. Most patients selected autografts for their ACL reconstruction. However, patients who performed significant Internet-based research preferred allografts. These results suggest that surgeons must acknowledge the Internet's influence in their patient populations. It is important to be aware of the magnitude and quality of the health care information available on the Internet due to its significant influence on patient graft decisions in ACL reconstruction surgery.