Shown is the cartilage shoe in vivo 5 years after implantation. The arrow indicates the cartilage shoe holding the total ossicular replacement prosthesis.
A, Shown are the mean air conduction (AC) and bone conduction (BC) hearing thresholds of all 42 ears. B, Shown are the numbers of patients relative to their pure-tone average air-bone gap (PTA-ABG) before surgery, 3 weeks after surgery (early follow-up), and 6.8 years after surgery (late follow-up).
eFigure 1A. Hearing Thresholds of Revision Tympanoplasties
eFigure 1B. Hearing Thresholds of Primary Tympanoplasties
eFigure 1C. Hearing Results, Revision Tympanoplasties & Primary Tympanoplasties
eFigure 2A. Hearing Thresholds of Canal Wall Down Procedures
eFigure 2B. Hearing Thresholds of Intact Canal Wall Tympanoplasties
eFigure 2C. Hearing Results, Canal Wall Down & Canal Wall Up Procedures
eFigure 3A. Hearing Thresholds of Own Tympanoplasties Needing Revision
eFigure 3B. Hearing Thresholds of Own Tympanoplasties Without Revision
eFigure 3C. Hearing Results, Revisions Following Tympanoplasty in Our Department & No Revision Surgery Following Surgery in Our Department
eFigure 4. Statistical Analysis and PTA-ABG of Each Patient Group
Customize your JAMA Network experience by selecting one or more topics from the list below.
Gostian A, Kouamé J, Bremke M, Ortmann M, Hüttenbrink K, Beutner D. Long-term Results of the Cartilage Shoe Technique to Anchor a Titanium Total Ossicular Replacement Prosthesis on the Stapes Footplate After Type III Tympanoplasty. JAMA Otolaryngol Head Neck Surg. 2016;142(11):1094–1099. doi:10.1001/jamaoto.2016.2118
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Does the cartilage shoe technique of anchoring a total ossicular replacement prosthesis on the footplate yield good long-term outcomes?
This retrospective medical record review shows that the air-bone gap could be closed to 30 dB or less in 76% of the patients after total ossiculoplasty. Intraoperative findings confirmed a firm connection of the prosthesis to the cartilage shoe.
The cartilage shoe technique is a reliable method with good functional long-term results for total ossiculoplasty.
Multiple techniques for a more secure and stable anchoring of the foot of a total ossicular replacement prosthesis (TORP) on the footplate have been proposed. To address this issue, a technique was developed that fits a cartilage shoe into the oval niche to guide the base of the prosthesis into a preformed central perforation.
To evaluate the long-term results of the cartilage shoe technique used to anchor a titanium TORP on the stapes footplate in patients after type III tympanoplasty.
Design, Setting, and Participants
Medical record review of total ossiculoplasties at a single center. The study included all patients who had undergone a tympanoplasty using a TORP between January 1, 2004, and December 31, 2008, at the Department of Otorhinolaryngology–Head and Neck Surgery, University of Cologne, Cologne, Germany.
Main Outcomes and Measures
Hearing thresholds were determined by a 4-frequency (500-4000 Hz) pure-tone average air-bone gap (PTA-ABG). Intraoperative findings from revision surgery and second-look operations are reported.
Forty-two ears were eligible for follow-up examination, averaging 6.8 years (range, 4.8-9.1 years) after surgery, that comprised otoscopy and audiometry. The mean age of 22 women and 20 men was 42.8 years (age range, 6-78 years). The overall PTA-ABG decreased from a mean (SD) of 33.0 (8.4) decibels (dB) before surgery to a mean (SD) of 22.0 (10.1) dB after surgery (P ≤ .001, η = 0.402). Before surgery, 64% (27 of 42) of the patients had a PTA-ABG exceeding 30 dB, which was 30 dB or less in 76% (32 of 42) of the patients after surgery. After canal wall down (n = 18) and tympanoplasties with intact canal wall (n = 24), the PTA-ABG was reduced from a mean (SD) of 33.0 (8.9) dB to 24.6 (11.2) dB (P = .01, η = 0.271) and from a mean (SD) of 32.0 (7.3) dB to 19.6 (9.2) dB (P ≤ .001, η = 0.511), respectively. For transmeatal revision ossiculoplasty (n = 33), the PTA-ABG improved from a mean (SD) of 32.0 (8.6) dB to 21.0 (10.2) dB (P ≤ .001, η = 0.389), similar to primary ossiculoplasty (n = 9), with a mean (SD) of 33.0 (5.5) dB PTA-ABG before surgery and a mean (SD) of 21.0 (9.8) dB PTA-ABG after surgery (P = .005, η = 0.478). Intraoperative findings from revision surgery (n = 11) during the study period and second-look procedures (n = 7) showed that the cartilage shoe prevented a lift off the footplate in all but 2 patients, even in the presence of recurrent cholesteatoma.
Conclusions and Relevance
The use of the cartilage shoe that anchors a titanium TORP on the stapes footplate can be advocated for good and reliable long-term results after total ossiculoplasty.
Modern reconstructive middle ear surgery offers a variety of industrially produced prostheses for restoration of the ossicular chain. Such implants have to be biocompatible (ie, they must produce no adverse reactions of body tissue). Implants that intend to restore a defect in the ossicular chain relative to hearing rehabilitation require a firm and stable connection to the remnants of the ossicular chain. This requirement is even more demanding for a type III tympanoplasty with missing stapedial arch and incus. In that case, the required total ossicular replacement prosthesis (TORP) placed solely on the stapes footplate is held by water adhesion forces only. Postoperative hematomas, scar formation, and displacements of the tympanic membrane due to changes in ambient static air pressure are potential risk factors for postoperative dislocation of the prosthesis to the bony annulus of the oval window, even in successfully implanted middle ear prostheses. This situation is associated with distinct conductive hearing loss up to the so-called transmission block of 50 to 60 decibels (dB).1 Thus, postoperative displacement of implanted middle ear prostheses is a major cause of unsatisfactory hearing results.2 Therefore, multiple techniques for a more secure and stable anchoring of the foot of the prosthesis on the footplate have been proposed.3-7 To address this issue, Hüttenbrink et al8 developed a technique that fits a cartilage shoe into the oval niche to guide the base of the prosthesis into a preformed central perforation. In vitro evaluation demonstrated a reliable and unimpaired sound transfer of this technique. In addition, encouraging clinical results are available up to a mean of 21.7 months after surgery.8-10 The objective of our study was to evaluate the long-term results of a cartilage shoe technique used to anchor a titanium TORP on the stapes footplate in patients undergoing type III tympanoplasty, which is routinely performed using the cartilage shoe technique.
This retrospective medical record study included all patients who had undergone a tympanoplasty using a TORP between January 1, 2004, and December 31, 2008, at the Department of Otorhinolaryngology–Head and Neck Surgery, University of Cologne, Cologne, Germany. Inclusion criteria were as follows: (1) patients who underwent tympanoplasty with ossicular chain reconstruction using a TORP with the cartilage shoe on an intact and mobile footplate of the stapes and (2) patients who had at least 1 audiogram before surgery and at follow-up. A total of 107 patients were evaluated according to the operation theater logbooks and medical record reviews. Care was taken (by means of directories, general practitioners, etc) to identify addresses of patients whose contact information was out of date. Accordingly, 42 of 62 eligible patients responded to our invitation for a follow-up examination. The presented results are derived from complete data sets of these 42 patients (ie, audiometric results from all 3 investigation periods and clinical examination at the last follow-up). The clinical follow-up examination comprised a medical history of the course after the initial surgery (eg, if and when additional operations had been performed) and microscopic otoscopy. Hearing thresholds were determined by pure-tone audiometry with a 4-frequency (500, 1000, 2000, and 4000 Hz) pure-tone average air-bone gap (PTA-ABG). In addition, audiometric data from the 42 patients 3 weeks after surgery (ie, at the time of dressing removal) was obtained from medical record review only. Herein, we refer to the time of dressing removal 3 weeks after surgery as early follow-up and the clinical follow-up, averaging 6.8 years after surgery, as late follow-up. The hearing outcomes at late follow-up were compared with the results at early follow-up and with the preoperative data. For analysis of the follow-up date (comparing the preoperative results with the early and late follow-up), a repeated-measures analysis of variance (ANOVA) was performed. Second, paired t tests were performed as post hoc tests for comparison of the PTA-ABGs. All P values were Greenhouse-Geisser corrected in case of nonsphericity.11 Statistical calculations were performed with a software program (SPSS, version 21.0; IBM). P < .05 was considered statistically significant.
In addition, effect sizes were computed for the results of the repeated-measures ANOVAs using the following formula to calculate partial η2p:
where SS indicates the sum of squares. The effect sizes displayed can be interpreted as small if they are less than 0.06, medium if they range from 0.06 to 0.14, and strong if they exceed 0.14. Effect sizes (r values) for the post hoc t tests were calculated as follows:
where t indicates the t statistic.11 Effect sizes for paired t tests are interpreted as small if r = 0.10, medium if r = 0.30, and large if r≥0.50.
This study was approved by the Ethics Committee of the Medical Faculty of the University of Cologne, Cologne, Germany, in accord with German law and the Declaration of Helsinki. Written informed consent was obtained from all patients enrolled.
Forty-two ears of 42 patients were eligible for the follow-up investigation. The mean follow-up time was 6.8 years (range, 4.8-9.1 years). The mean age of 22 women and 20 men was 42.8 years (age range, 6-78 years). Tympanoplasty was performed in 23 right ears and 19 left ears using general anesthesia via the retroauricular approach. The Table lists the characteristics of the patients. Thirty-three patients had undergone previous surgical procedures, with a maximum of 8 prior tympanoplasties in 1 patient. We performed a canal wall down (CWD) procedure in 7 ears and revised 11 ears with a previously performed CWD procedure. All radical cavities were partially obliterated according to the technique by Beutner et al.12 Eleven ears needed revision surgery after tympanoplasty in our department due to recurrent cholesteatoma after a mean of 3.4 years (range, 0.5-6.7 years) after surgery. Nine of these patients had undergone previous surgery elsewhere.
Ossiculoplasty and eradication of disease were performed in the same step in all ears. In case of extensive cholesteatoma and an intact canal wall procedure, a second-look operation was performed a mean of 0.6 years (range, 0.4-0.8 years) after surgery based on the judgment of the surgeon during surgery. All tympanoplasties performed in our department used the cartilage shoe anchoring a TORP on the footplate. The hearing thresholds before surgery and 3 weeks after surgery refer to the initial surgical procedure in our department, whereas the outcomes after revision surgery in our department contributed to the results at late follow-up. All cartilage used during surgery was harvested from the cavum conchae of the same ear and prepared using available equipment (Precise Cartilage Knife Set; Heinz Kurz Company). The surgical technique involving the cartilage shoe was performed according to the detailed description by Beutner et al.10 In all cases, the prostheses were covered with a piece of cartilage approximately 0.5 mm thick (ie, a piece of cartilage was always placed laterally on the head of the prosthesis underneath the tympanic membrane).
At follow-up examinations, all eardrums were intact except for one that showed a small defect of approximately 0.5 × 1 mm in the anterosuperior quadrant of the tympanic membrane. No ear revealed any evidence or suspicion of cholesteatoma. Furthermore, all ears were dry without any signs of inflammation or granulation. None of the patients reported any additional surgery performed elsewhere since the last surgical procedure in our department.
In 1 of our 11 revision tympanoplasties, cholesteatoma led to conductive hearing loss because it had lifted up the TORP with the cartilage shoe still attached to it. In all other surgical procedures, the contact of the TORP to the cartilage shoe remained stable, despite the existence of recurrent cholesteatoma.
In 7 patients, we performed second-look operations. During surgery, we found in one case that the head of the prosthesis was bent anteriorly, while the base of the prosthesis remained in the center of the cartilage shoe on the stapes footplate. In another case, the TORP (4.5 mm) had been lifted off and separated from the cartilage shoe. However, the cartilage shoe remained stable in the oval niche, allowing for placement of a longer prosthesis (5.5 mm). All other second-look operations revealed no evidence of residual cholesteatoma, with a stable and secure fit of the TORP into the cartilage shoe in the oval window niche. In Figure 1, the intraoperative finding of a cartilage shoe holding the base of the TORP 5 years after implantation is shown.
Figure 2A shows the mean hearing results in all 42 patients. The overall PTA-ABG decreased from a mean (SD) of 33.0 (8.4) dB before surgery to a mean (SD) of 23.0 (8.7) dB 3 weeks after surgery and to a mean (SD) of 22.0 (10.1) dB at the late follow-up examination. The repeated-measures ANOVA found a significant main effect follow-up date (F1.6,64.4 = 27.5, P ≤ .001, η = 0.402).
Before surgery, no ear had a PTA-ABG less than 10 dB. However, the PTA-ABG ranged from 10 to 30 dB in 15 ears (36%) and exceeded 30 dB in 27 ears (64%). The corresponding numbers 6.8 years after surgery were 5 (12%), 27 (64%), and 10 (24%), respectively (Figure 2B).
In 33 ears that had undergone previous surgery, the PTA-ABG decreased from a mean (SD) of 32.0 (8.6) dB before surgery to a mean (SD) of 22.5 (9.2) dB at the time of dressing removal and to a mean (SD) of 21.0 (10.2) dB at late follow-up (F1.5,48.2 = 20.4, P ≤ .001, η = 0.389). The corresponding numbers in 9 cases of primary surgery were means (SDs) of 33.0 (5.5) dB, 24.1 (9.6) dB, and 21.0 (9.8) dB, respectively (F2,16 = 7.3, P = .005, η = 0.478) (eFigure 1A and B in the Supplement).
Before surgery, all 33 of 33 (100%) revision tympanoplasties had a PTA-ABG exceeding 10 dB, while 20 of 33 ears (61%) had a PTA-ABG of 31 dB or less. At the time of dressing removal, the PTA-ABG had dropped below 10 dB in 3 of 33 ears (9%) and was still present in 5 of 33 ears (15%) 6.8 years after surgery. Accordingly, the finding of a PTA-ABG exceeding 31 dB was reduced from 20 of 33 ears (61%) before surgery to 8 of 33 ears (24%) at late follow-up. All 9 primary tympanoplasties had a PTA-ABG exceeding 20 dB before surgery. At late follow-up, the PTA-ABG did not exceed 20 dB in 6 of 9 ears (67%).
eFigure 1C in the Supplement shows the number of patients relative to their preoperative and postoperative PTA-ABG. We performed 7 primary CWD procedures and revised 11 previously performed CWD procedures. Therefore, 18 patients with CWD procedures were eligible for evaluation. Their hearing thresholds are shown in eFigure 2A in the Supplement. The PTA-ABG of 18 CWD procedures decreased from a mean (SD) of 33.0 (8.9) dB before surgery to a mean (SD) of 26.2 (10.4) dB 3 weeks after surgery and to a mean (SD) of 24.6 (11.2) dB at late follow-up (F1.4,23.5 = 6.3, P = .01, η = 0.271). For the 24 intact canal wall tympanoplasties procedures, the corresponding numbers were means (SDs) of 32.0 (7.3) dB, 20.5 (6.4) dB, and 19.6 (9.2) dB, respectively (F1.4,23.4 = 6.3, P ≤ .001, η = 0.511) (eFigure 2A and B in the Supplement).
At late follow-up, 2 of 18 ears (11%) with CWD procedures had a PTA-ABG less than 11 dB, in contrast to the preoperative state and to the time of dressing removal, when all ears had air-bone gaps exceeding 10 dB. In addition, the number of patients with a PTA-ABG of 11 to 20 dB increased from 1 of 18 ears (6%) before surgery to 6 of 18 ears (33%) 6.5 years after surgery.
In 24 cases in which the posterior wall of the external ear canal was left intact, only 1 ear (4%) had a PTA-ABG less than 21 dB before surgery. At late follow-up, 15 of 24 ears (63%) had a PTA-ABG less than 21 dB, with 4 of 24 ears (17%) exhibiting an air-bone gap less than 11 dB.
eFigure 2C in the Supplement shows the number of ears according to their PTA-ABG before surgery and 3 weeks and 6.8 years after the dressing removal. Eleven ears needed revision tympanoplasty with a cartilage shoe in our department. eFigure 3A in the Supplement shows the hearing outcomes of our revision tympanoplasties, which are of the same magnitude as the functional outcomes of 31 ears without the need for revision surgery after surgery in our department (eFigure 3B in the Supplement). The mean PTA-ABG in this subgroup decreased from a mean (SD) of 33.5 (6.6) dB to a mean (SD) of 24.1 (9.6) dB 3 weeks after surgery and improved further to a mean (SD) of 22.4 (10.6) dB at late follow-up (F2,20 = 8.19, P = .003, η = 0.450).
In cases without the need for revision surgery after tympanoplasty in our department, the PTA-ABG dropped from a mean (SD) of 32.1 (8.5) dB to a mean (SD) of 22.6 (8.6) dB at the time of dressing removal 3 weeks after surgery. In the long term, the hearing outcomes remained stable, with a mean (SD) PTA-ABG at late follow-up of 21.5 (10.5) (F1.6,48.1 = 18.9, P ≤ .001, η = 0.387).
eFigure 3C in the Supplement shows the respective hearing levels of the patients with and without revision surgery after tympanoplasty in our department. In the patient group with revision tympanoplasty (n = 11), 1 ear (9%) had a PTA-ABG less than 11 dB at late follow-up. In contrast, before surgery and 3 weeks after surgery, all patients had a PTA-ABG exceeding 10 dB. In addition, the number of ears with a PTA-ABG of 11 to 20 dB increased from 1 ear before surgery to 4 ears at late follow-up. Accordingly, a PTA-ABG between 21 and 30 dB and exceeding 31 dB was present before surgery in 5 of 11 patients (45%) and 6 of 11 patients (55%), respectively. At late follow-up, the respective numbers were reduced to 3 of 11 patients (27%).
Before surgery, only 2 of 33 ears (6%) that we had operated on without the need for revision tympanoplasty had a PTA-ABG less than 21 dB. At 6.8 years after surgery in our department, the number increased to 18 of 33 ears (55%). In the same period, the number of ears with a PTA-ABG of at least 31 dB was reduced from 20 of 33 (61%) to 7 of 33 (21%).
Analysis of the PTA-ABG in each subgroup of patients revealed a statistically significant improvement in hearing outcome, together with large effect sizes from the preoperative state to the early postoperative follow-up. These subgroups included all patients (t41 = 5.34, P ≤ .001, r = 0.64), revision surgical procedures (t32 = 4.66, P ≤ .001, r = 0.64), primary surgical procedures (t8 = 2.62, P = .03, r = 0.68), CWD (t17 = 3.01, P = .008, r = 0.59), intact canal wall procedures (t23 = 5.54, P ≤ .001, r = 0.76), own revision surgical procedures (t10 = 2.63, P = .03, r = 0.64), and ears operated on only once (t30 = 4.57, P ≤ .001, r = 0.64). In contrast, comparison of the hearing outcomes from early to late follow-up lacked statistically significant differences. These data for the subgroups were as follows: all patients (t41 = 1.13, P = .26, r = 0.17), revision surgical procedures (t32 = 0.58, P = .57, r = 0.10), primary surgical procedures (t8 = 1.25, P = .25, r = 0.40), CWD (t17 = 1.10, P = .29, r = 0.25), intact canal wall procedures (t23 = 0.59, P = .56, r = 0.12), own revision surgical procedures (t10 = 0.94, P = .37, r = 0.28), and ears operated on only once (t30 = 0.79, P = .43, r = 0.14). eFigure 4 in the Supplement shows the results of the repeated-measures ANOVAs for each subgroup.
This study evaluated the long-term hearing results after type III tympanoplasty using a TORP that was secured on the stapes footplate by a cartilage shoe as described by Beutner et al.10 Given the poor middle ear condition of the patient cohort herein, the use of a cartilage shoe yielded good functional and stable long-term results. In addition, the hearing outcomes measured 3 weeks after surgery showed no significant differences compared with late follow-up almost 7 years after surgery. Intraoperative findings during revision and second-look surgery confirmed a firm connection of the cartilage shoe to the prosthesis shaft.
Stable and reliable placement of the titanium prosthesis and the remnants of the ossicular chain is a prerequisite for successful hearing results in modern tympanoplasty.13 Placement of a TORP on the stapes footplate is a challenging operative procedure for the otosurgeon. Firm placement during surgery may still be altered once the ear has been closed for various reasons, such as hematoma, scar formation, middle ear effusion, atmospheric pressure changes, and recurrence of disease. Therefore, a variety of techniques aiming to secure the prosthesis shaft on the footplate have been advocated using connective tissue or cartilage.14,15 The cartilage shoe technique fits the tailored shoe into the oval niche to fit securely between the facial nerve and the promontory. In addition, the surrounding bone structure prevents dislocation of the cartilage shoe from the stapes footplate. Experimental evaluation of the presented cartilage shoe technique showed that the mass of the cartilage shoe and the friction of the foot of the prosthesis in the central perforation of the cartilage shoe did not impair sound transmission initially. However, sound transmission was impaired by no more than 5 dB by gluing the cartilage shoe to the footplate, which mimicked its stable scarred ingrowth after implantation in vivo.8 Clinically, the audiometric results presented herein almost 7 years after surgery, with a reduction in the overall PTA-ABG from a mean (SD) of 33.0 (8.4) dB before surgery to a mean (SD) of 22.0 (10.1) dB after surgery, are comparable to those reported in the literature with other techniques and prostheses.16-22 With regard to the cartilage shoe, the positive hearing results can also be attributed to stabilization of the TORP in the center of the footplate, which is the optimal site for energy transfer.23
Fayad et al9 presented the outcomes of columella-type tympanoplasties using a TORP with and without a cartilage shoe, comparable to the results of our study. However, compared with their study, the patient cohort in our study had a greater degree of middle ear destruction, as noted by the higher percentages of CWD procedures (43% [18 of 42] vs 12.2%) and revision surgery (79% [33 of 42] vs 48.7%), which have been reported to be associated with a poorer functional outcome.24-26 In addition, in our department, we routinely perform ossicular reconstruction simultaneously in one step with eradication of the disease. In contrast, 43.2% of the cases in the study by Fayad et al9 were planned second-stage procedures, which can be expected to yield better functional results due to a stable middle ear and intact tympanic membrane at the time of ossicular reconstruction.
With regard to hearing results, satisfactory long-term outcomes have been reported for porous polyethylene (Plastipore; Medtronic) total ossicular reconstruction prostheses.27 Berenholz et al27 achieved closure of the air-bone gap to 20 dB or less in 60% of their patients a mean of 4.3 years after total ossiculoplasty, with a mean PTA-ABG of 20.2 dB, which is comparable to our study. Considering the costs of total ossiculoplasty, the Plastipore TORP is available at a lower retail price than the titanium TORP (approximately 100€ [US $110] vs 155€ [US $171], respectively).
The results herein demonstrate that the cartilage shoe provides stability to the base of the prosthesis in the horizontal plane. Intraoperative findings during revision surgery showed that the cartilage shoe was anchored securely on the footplate in the oval niche, although the prosthesis had been bent anteriorly in one case. However, this technique does not add stability in the vertical plane. Anchoring of the TORP with the cartilage shoe does not completely prevent the prosthesis with or without the cartilage shoe to be lifted off the footplate and consequently lead to conductive hearing loss.10 Nevertheless, limited vertical mobility of the foot of the prosthesis in the central perforation of the cartilaginous shoe can be assumed depending on the thickness of the shoe and the length of the prosthesis.
The presented data demonstrate our postsurgical results almost 7 years after applying a single reconstruction technique with a titanium prosthesis. However, we are aware of the issues associated with long-term investigations and loss to follow-up. Despite great effort, we gathered contact information from only 58% (62 of 107) of the initially eligible patients. Likely reasons are the mobility in Germany’s highly populated Rhein-Ruhr area, with approximately 10 million inhabitants, and the large area from which patients are being referred to our department. Despite the limited number of patients eligible for evaluation, our results still provide useful additional information on long-term outcomes of the cartilage shoe technique of securing a titanium TORP on the stapes footplate.
Ossiculoplasty involving a TORP placed on the bare footplate is challenging. The cartilage shoe technique can be recommended for secure, standardized, and stable placement of the TORP in the center of the stapes footplate prosthesis, which offers good morphologic and functional long-term results.
Accepted for Publication: June 15, 2016.
Corresponding Author: Antoniu-Oreste Gostian, MD, Department of Otorhinolaryngology–Head and Neck Surgery, Medical Faculty, University of Cologne, 50924 Cologne, Germany (firstname.lastname@example.org).
Published Online: August 18, 2016. doi:10.1001/jamaoto.2016.2118.
Author Contributions: Dr Gostian had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Gostian, Bremke, Beutner.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Gostian, Kouamé.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Gostian, Kouamé, Ortmann.
Administrative, technical, or material support: Gostian, Bremke, Beutner.
Study supervision: Gostian, Hüttenbrink, Beutner.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest, and none were reported.
Create a personal account or sign in to: