Disclosed Industry Funding Does Not Increase Positive Outcomes in Studies on Suture Button Fixation for Ankle Syndesmotic Injuries: A Systematic Review

Abstract

Background:

Industry funding can increase the likelihood of positive outcomes. This study sought to investigate whether industry funding or conflicts of interest (COI) for studies investigating suture button fixation devices for ankle syndesmosis injuries influenced outcomes.

Methods:

PubMed, Scopus, and Embase were queried for studies investigating suture button fixation devices for the ankle. Included studies investigated suture button fixation devices for the ankle and reported funding or COI. Excluded studies were systematic reviews, meta-analyses, and case reports. Outcomes were categorized into “positive” if the null hypothesis was rejected or outcomes favored the implant, “neutral” if the null hypothesis was confirmed, or “negative” if the result was unfavorable toward the implant. Studies were grouped into industry-funded, nonfunded, and “other” funding, including specialty societies, National Institutes of Health (NIH), and grants. Studies were also grouped by the presence or absence of COI. Chi-squared test was used to test for significance defined as P <.05.

Results:

A total of 112 studies were included for analysis. Of these, 19 (17%) were industry-funded, 21 (19%) had other funding sources, and 72 (64%) were nonfunded. The proportion of studies with positive outcomes was 48% in industry-funded studies, 63% in studies with other funding sources, and 65% in nonfunded studies. There was no significant association between funding source and outcome (P = .063). A COI was present in 42 studies (37.5%), and no COI was present in 70 studies (62.5%). The proportion of positive studies in those with a COI was 52%, whereas in studies without a COI, it was 67%. Positive studies were significantly associated with studies without COI (P = .003).

Conclusion:

Industry funding and COI was not found to be associated with an increased likelihood of positive outcome reporting in studies on suture-button fixation for ankle syndesmostic injuries; we found in this review that positive outcomes are more likely in studies without COI.

Introduction

Ankle syndesmotic injuries, which disrupt the ligamentous structures stabilizing the distal tibiofibular joint, pose a significant challenge in orthopaedic practice due to their potential for long-term complications, including joint instability and posttraumatic arthritis. ¹⁹ Among the various treatment modalities available, suture button fixation has gained prominence as an alternative to traditional methods. The purported advantages of suture button systems include dynamic stabilization, reduced risk of implant failure, and earlier return to function. ¹²⁵ These features have positioned suture buttons as a popular choice, particularly in complex cases, and have sparked substantial interest in their clinical outcomes.^80,126

With the growing adoption of suture button fixation, industry funding has increasingly supported research in this area, as device manufacturers attempt to obtain market share of a developing space. Although industry funding is critical for driving innovation and enabling large-scale clinical trials, it raises concerns about potential biases in the design, conduct, and reporting of studies. Evidence from other areas of orthopaedic surgery suggests that industry-funded research is more likely to report favorable outcomes, potentially skewing the clinical perception of implant performance.^3,35,66,76 Furthermore, industry funding can cause biases even within high-level evidence studies, like randomized controlled trials (RCTs). Industry funding has been shown to affect methodology in RCTs and is associated with discrepancies between primary outcomes of the RCT registration and its publication.^31,62,67 Given the commercial and clinical implications of suture button fixation devices, it is essential to critically evaluate whether funding sources influence reported outcomes and complication rates.

This study seeks to systematically analyze the impact of industry funding on the findings of studies investigating suture button fixation in ankle syndesmotic injuries. By examining trends in funding patterns and reported outcomes, this study aims to assess the extent of potential bias in the literature and provide clinicians with a clearer understanding of the evidence base supporting suture button systems. We hypothesized that studies with industry funding would be more likely to report positive outcomes.

Methods

Literature Screening

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, PubMed, Scopus, and Web of Science were queried for studies investigating suture-button fixation devices for the ankle syndesmosis. The search strategy was as follows: (“ankle” OR “syndesmo*” OR “tibiofibular”) AND (“suture button” OR “button” OR “tightrope”). Included studies were those that were investigating suture-button fixation for ankle syndesmosis, reported on the presence or absence of funding sources or conflicts of interest (COI), and were published in English in peer-reviewed journals. Exclusion criteria were: systematic reviews and meta-analyses, case reports, abstracts, and studies with no full text. Any study that failed to mention funding information or conflicts of interest was excluded.

This study was not registered with PROSPERO in advance. A review protocol is available upon request.

Screening and Grouping

Studies were compiled in Covidence (Melbourne, Australia). Two authors (C.L.F. and E.H.L) independently screened studies by title and abstract before 2 additional authors (A.A.J. and S.C.T) screened studies by full text. Disputes were resolved by a third author (A.K.S.). Two authors (A.A.J. and S.C.T) categorized studies by funding. These groups included “nonfunded,” if the study explicitly indicated that no funding was used; “industry-funded,” which included studies using funding, grants, and donated materials from industry-affiliated companies; and “other funding,” which included studies using grants or support from specialty societies, government, and institutional. Studies were also grouped by the presence or absence of COI.

Two authors (C.L.F. and E.H.L) independently categorized studies into groups based on outcome. Studies were deemed to have “positive” outcomes if the results significantly favored the implant by, for example, demonstrating better clinical or patient-reported outcomes, performing better in biomechanical studies, or showing cost-effectiveness compared with alternative options. Studies were deemed “neutral” if the null hypothesis failed to be rejected or if there were multiple endpoints and some favored while others disfavored the implant, or the endpoint result could be deemed as neither favorable nor unfavorable for the implant. Studies with “negative” outcomes were those whose findings did not favor the implant, if the study recommended against its use, if it performed worse than alternatives in biomechanical studies, or it was found to be cost-ineffective compared with alternative management options.

Quality Assessment

A quality assessment tool was used to evaluate the quality of each study included. For clinical studies, the Methodological Index for Non-Randomized Studies (MINORS) was used. ¹⁰⁴ For comparative clinical studies, the maximum MINORS score is 24, and for noncomparative clinical studies, the maximum MINORS score is 16. For cadaver studies, the Quality Appraisal for Cadaveric Studies (QUACS) scale was used. ¹¹⁹ The maximum QUACS score is 13.

Data Analysis

All data was recorded descriptively. RStudio (Boston, MA) was used for all statistical analyses. Fisher exact test was used to test for association between outcome type and funding category. Statistical significance was set at P < .05.

Results

Screening

Initial search results yielded 1246 studies, from which 618 duplicates were removed by Covidence. During the first phase of the screen, 486 studies were removed after being found to not fit inclusion criteria based on title and abstract. These included systematic reviews, case reports, use of suture-button devices for other anatomical regions, etc. In the full-text screen, 26 studies were removed. The full text could not be located for 3, funding and COI were not reported in 1, 13 studies were not in English, 4 were duplicate studies, and 5 were of the wrong study design. A total of 112 studies were included for analysis (Figure 1).^{1,2,4,5,8,9,11 -18,20 -22,24 -26,28 -30,32 -34,36 -59,61,63 -65,68 -75,77 -100,102,103,105 -118,120 -124,126,128 -132}

Figure 1.

PRISMA diagram demonstrating search results.

Study Characteristics

The included studies were published between 2005 and 2024. A majority (54%) of these studies were published in or after 2020. Of the 112 studies, 8 (7.1%) were of level of evidence (LOE) I, 10 (8.9%) were of LOE II, 18 (16.1%) were of LOE III, 51 (45.5%) were of LOE IV, and 25 (22.3%) were of LOE V. Of the 73 clinical studies that underwent quality assessment using MINORS, 38 (52.1%) were comparative and 35 (47.9%) were noncomparative. The average MINORS score for comparative studies was 17.8 ± 3.8 and the average MINORS score for noncomparative studies was 9.8 ± 2.4. The average QUACS score for the 34 cadaveric studies was 12.9 ± 0.9. Five studies could not be assessed using MINORS or QUACS as they were cost analysis, cost model, and computer model studies.

Outcomes and Funding

Industry funding was used in 19 studies (17%),^{13,16,25,30,32,38,57,64,75,82,84,85,95,97,99,100,109,122,123} other sources of funding were used in 21 studies (19%),^{5,9,11,26,42,44,46 -48,54,58,59,68,71,72,106,108,114,118,131,132} and no funding was used in 72 studies (64%).^{1,2,4,9,12,14,15,17,18,20 -22,24,26,28,29,33,34,36,37,39 -41,43,45,49 -51,53,55,56,61,63,65,69,70,73,74,77 -86,81,86 -94,96,98,102,103,105,108,111 -113,115 -117,120,121,124,126,128 -130} Of all the 112 studies included, 69 (61.6%) demonstrated positive outcomes, 26 (23.2%) demonstrated neutral outcomes, and 17 (15.2%) demonstrated negative outcomes. When broken down by funding type, of the industry-funded studies, 10 (52%) were positive, 7 (37%) were negative, and 2 (11%) were neutral. Studies with other sources of funding were positive in 12 studies (57%), neutral in 5 studies (24%), and negative in 4 studies (19%). For nonfunded studies, 47 (65%) had positive results, 18 (25%) had neutral results, and 7 (10%) had negative results. Outcomes reporting was not significantly associated with study funding type (P = .063) (Table 1).

Table 1.

χ² 3 × 3 contingency table of outcomes grading by funding type.

	Positive Outcome	Negative Outcome	Neutral Outcome
Industry funding	10	7	2
Nonfunded	47	7	18
Other funding	12	4	5
P value			.063

The presence of at least 1 COI was identified in 42 studies (37.5%)^{5,13,16,20,26,29,30,32,38,39,49,51,54,57,58,64,70,75,79 -86,88,94,95,97,99,100,105,107 -109,112,117,118,122,123,131} and no COI was present in 70 studies (62.5%).^{1,2,4,8,9,11,12,14,15,17,18,21,22,24,28,34,36,37,40 -48,50,52,53,55,56,59,61,63,65,68,69,71 -74,75,78,87,89 -93,96,98,102,103,106,110,111,113 -116,120,121,124,126,128 -132} Positive outcomes were seen in 22 studies (52%) with a COI vs in 47 studies (67%) without a COI. Negative outcomes were seen in 13 studies (31%) with a COI and in only 5 studies (7%) without COI. Outcome reporting was significantly associated with the presence or absence of COI (P = .003) (Table 2). Positive outcomes were more likely among studies without COI, and negative outcomes were more likely among studies with COI.

Table 2.

χ² 3 × 2 Contingency Table Comparing Study Outcomes by the Presence or Absence of COI.

	Positive Outcomes	Neutral Outcomes	Negative Outcomes
COI present	22	7	13
COI absent	47	18	5
P value			.003

Discussion

This study demonstrates that industry funding shows no association with outcomes and that the presence of COI has the opposite association of what would be expected, with these studies being more likely to report negative or neutral outcomes than studies without funding or COI. The majority of studies on suture button fixation received no funding, whereas less than a quarter were industry-funded. Outcomes that favored suture button fixation devices were found in more than half of the included studies.

The cost of bringing a new medical device to market is one of the driving forces behind industry bias in research. Rigid fixation using screws is the historical option for managing injuries to the ankle syndesmosis. Suture button fixation devices were introduced to the landscape more recently, with the first receiving Food and Drug Administration (FDA) approval in 2005. ⁷ For the industry, developing new medical devices and gaining FDA approval can be a long and expensive process. Sertkaya et al ¹⁰¹ conducted an economic analysis to identify how costly the FDA approval process can be. When accounting for failed studies and the cost of capital, the mean price for bringing a novel therapeutic complex device to market is more than $500 million. An expedited approval process known as the 510(k) premarket notification exists, which can be used for devices substantially equivalent to an existing device. Yang et al ¹²⁷ found that for orthopaedic devices approved through the 510(k) premarket notification, as multiple suture-button devices received,^7,27 the average amount of time for industry to recoup investments was more than 2 years. Given the resources required to develop and bring new devices to market, it is understandable that industry companies would fund research to help demonstrate the efficacy and safety of their product.

Industry sponsorship in research does not only benefit the companies when leading to the publication of favorable findings, but it also helps fuel higher-impact studies. In orthopaedic research, industry-sponsored studies make up the majority of level I studies. ²³ Further, industry funding is associated with higher numbers of patients involved in research and multicontinental recruitment locations. ⁶⁰ Thus, industry funding is beneficial in supporting higher-powered studies with greater generalizability. However, this comes with the risk of industry bias, which has been well-established in orthopaedic research.^3,35,66 The findings of this study demonstrate that in research on dynamic suture-button fixation of the ankle syndesmosis, the presence of COI is instead associated with higher rates of neutral and negative outcomes. This may suggest an increased awareness of industry biases in orthopaedic research and an active effort to increase transparency among studies with these types of affiliations. Therefore, clinicians can potentially use the findings of research on these devices with less concern about this bias affecting their clinical decision making. In bolstering high-impact research, industry funding may also strengthen the literature on suture-button fixation.

The majority of research on suture-button fixation for the ankle demonstrated favorable results. These positive outcomes were not associated with funding sources. This is in contrast with research on shoulder arthroplasty and new technologies in hip and knee arthroplasty, for example, where industry funding is associated with favorable outcomes.^10,35 Importantly, funding bias is prevalent within orthopaedic literature more broadly. Dubin et al ²³ found that conflicts of interest were significantly associated with positive outcomes in more than 1000 studies in the top 3 high-impact orthopaedic journals. To accurately assess for funding bias, transparency in the reporting of funding and conflicts of interest is required. Anz et al ⁶ demonstrated that in clinical trials on biological treatments in orthopaedics, 49% of studies did not report study funding. Studies on suture-button fixation were excluded if they did not report on funding. Therefore, it was impossible to account for studies that may have lacked this transparency in this analysis. Hence, caution is always recommended when interpreting study results.

Strengths of this study include adherence to PRISMA guidelines, resulting in a comprehensive search strategy using multiple major literature databases. This study provides practical insights to both research and clinicians when evaluating the outcomes of suture button studies. This study is not without limitations. First, the sample sizes of the nonfunded and other funding groups were much smaller than the industry-funded group. This reduces the power to detect significant differences. Second, despite efforts to create a comprehensive search strategy, it is possible our search did not capture all studies that meet inclusion criteria. The studies included in this analysis were also quite heterogeneous. To provide an analysis generalizable to the entire ankle suture-button body of literature, we did not use inclusion or exclusion criteria that specified injury type, presence or absence of concomitant procedures, or endpoint measures. This study was also not able to stratify by amount of industry funding. For example, studies receiving payment from industry sources to carry out the study and studies that simply received donated devices were grouped together. We were also not able to account for how much money was contributed to each study, as this information is unavailable. Further, as a cross-sectional study, we are not able to identify causation, but only correlation. Future research into this topic could work to identify if industry funding in suture-button fixation research is associated with higher level of evidence research.

Conclusion

Funding type was not associated with outcomes but studies with at least 1 COI were significantly more likely to produce outcomes that were negative or neutral toward the suture-button construct for the ankle. The majority of the body of literature on this device demonstrates findings that are favorable toward suture-button devices. Clinicians using research to help inform their practice should always be concerned for the presence of biases within research.