Abstract
Background:
Treatment recommendations for trapeziometacarpal (TMC) arthrosis are highly variable from surgeon to surgeon. This study addressed the influence of viewing radiographs on a decision to offer surgery for TMC arthrosis.
Methods:
In an online survey, 92 hand surgeons viewed clinical scenarios and were asked if they would offer surgery to 30 patients with TMC arthrosis. Forty-two observers were randomly assigned to review clinical information alone and 50 to review clinical information as well as radiographs. The degree of limitation of daily activities, time since diagnosis, prior treatment, pain with grind, crepitation with grind, and metacarpal adduction with metacarpophalangeal hyperextension were randomized for each patient scenario to determine the influence of these factors on offers of surgery. A cross-classified binary logistic multilevel regression analysis identified factors associated with surgeon offer of surgery.
Results:
Surgeons were more likely to offer surgery when they viewed radiographs (42% vs. 32%, P = 0.01). Other factors associated variation in offer of surgery included greater limitation of daily activities, symptoms for a year, prior splint or injection, deformity of the metacarpophalangeal joint. Factors not associated included limb dominance, prominence of the TMC joint, crepitation with the grind test, and pinch and grip strength.
Conclusion:
Surgeons that view radiographs are more likely to offer surgery to people with TMC arthrosis. urgeons are also more likely to offer surgery when people do not adapt with time and nonoperative treatment. Given the notable influence of surgeon bias, and the potential for surgeon and patient impatience with the adaptation process, methods for increasing patient participation in the decision-making process merit additional attention and study.
Key Words: Decision-making, Interobserver variation, Osteoarthritis, Surgery, Trapeziometacarpal arthrosis, Treatment
Introduction
Treatment recommendations for trapeziometacarpal (TMC) arthrosis are highly variable from surgeon to surgeon. Spaans et al. (1) found fair interobserver agreement among eight surgeons on preferred treatment for 40 patients with TMC arthrosis based on radiographs, symptoms, and the Disabilities of the Arm, Shoulder, and Hand score. The greatest variability was amongst Eaton-Littler stages 2 and 3. Another retrospective study found a wide variety of practice variation for hand osteoarthritis amongst six hand surgeons (2). A survey of active members of the American Society for Surgery of the Hand found that 62% preferred trapeziectomy with ligament reconstruction and tendon intrposition (LRTI) for Eaton stage 3, despite the evidence that trapeziectomy alone is just as effective with fewer complications (3-5). Additional research is merited to determine the factors that account for surgeon-to-surgeon variation in treatment recommendations in order to aid efforts to reduce the influence of surgeon preferences and values in favour of the preferences and values of the patient.
This study randomized a large number of hand surgeons to give treatment recommendations based on randomly varied circumstances of patients with TMC arthrosis in order to address the primary null hypothesis that there are no factors associated with a recommendation for operative treatment for TMC arthrosis among patient factors, surgeon factors, and radiographic factors.
Materials and Methods
Two-hundred and eighty-three hand surgeons of the Science of Variation Group (SOVG) (6), a web-based collaborative, were invited between November 2014 and January 2015 to participate in an online survey about their preferred treatment method for presentations of patients with TMC arthrosis based on clinical information alone or on clinical information in combination with radiographs [Figure 1]. Hand surgeons that did not complete the survey within three weeks received a reminder email. Observers did not receive any incentives other than acknowledgement that they were part of the SOVG. There was no time limit to complete the survey, but we collected the online data after 8 weeks. This study was approved by the institutional review board (IRB).
Figure 1.
The scenario prior to randomization of the elements for the lateral (A), oblique (B), and Robert (C) radiographs depicted in the figure: A 62 year-old man has pain at the base of her (dominant/nondominant) thumb. The pain (does not limit/slightly limits/moderately limits/severely limits/makes impossible) daily activities. The patient was diagnosed with trapeziometacarpal arthrosis (now/ 3 months ago/ 1 year ago). The patient has (not treated it/ tried a splint/ tried a cortisone injection). On physical exam there (is / is not) fullness at the base of the thumb (so called “shoulder sign”), and (no/ mild/ moderate, severe/ extreme) pain and (no/ mild/ moderate/ substantial/ severe) crepitation with grind test. There is no adduction contracture and no MCP hyperextension. There (is / is not) an adduction contracture and MCP hyperextension. Pinch and grip strength are (good / poor)
Thirty consecutive patients diagnosed with TMC arthrosis were identified from a database. Inclusion criteria were (1) adult patients (≥18 years) and (2) adequate quality set of radiographs (Robert [true anteroposterior] view, true lateral view, and oblique view). Exclusion criteria were (1) pregnancy (IRB mandated) and (2) prior surgery for TMC arthrosis on the selected side. The diagnosis of TMC arthrosis was based on symptoms, physical examination and radiographic findings. Observers were randomly assigned on a 1:1 basis to choose their preferred method of treatment for the 30 patient presentations based on clinical information alone or on clinical information in combination with radiographs. Patients were presented in a random order.
Clinical information consisted of demographic information (sex, age, dominant or nondominant thumb), symptoms (how much the pain limited daily activities [not, slightly, moderately, severely, or impossible to perform activities]), previous diagnosis (no previous diagnosis, 3 months ago, or 1 year ago), previous treatment (none, splint, or splint and cortisone injection) and physical examination (fullness at the base of the thumb [‘shoulder sign’; yes or no], pain [none, mild, moderate, severe, or extreme] with the grind test, crepitation [none, mild, moderate, substantial, or severe] with the grind test, adduction contracture and metacarpophalangeal hyperextension [yes or no], and pinch and grip strength [poor or good]). The above-mentioned clinical information of all 30 patients was fictional, except for sex, age and radiographs which were extracted from 30 ‘real’ patients. All fictive patient characteristics were randomly assigned with a balanced distribution over the possible categories for each patient presentation every time an observer rated a patient [Figure 1]. This created a large random set of patients. Observers were unaware that the clinical information for patient presentations was both fictive and randomized.
Statistical analysis
We evaluated 15 parameters for the observer characteristics and 23 parameters for the patient characteristics in 2760 measurements (92 unique observers * 30 patients). Given the rule of thumb of a minimum of 10 independent measurements per parameter, we had sufficient power for our goal.
Due to the hierarchical data structure, a cross-classified binary logistic multilevel regression analysis was used to identify factors that determine when hand surgeons (observers) recommend surgery. Data was clustered both within patient presentations and observers since all observers rated all patients. Elements of the clinical information that were reported on 5-point Likert scales (e.g., pain with grind test) were considered categorical variables for analyses. First of all, each individual predictor variable (i.e., each patient or surgeon characteristic) was separately entered as a fixed effect into a regression model, which further only included random intercepts for patient presentation and observer, to examine the bivariable relationship. Variables with P<0.15 in this bivariable analysis, were included as candidate fixed effects in the final multivariable regression analysis, again including the random intercepts for patient presentation and observer. Variable selection by the backward stepwise approach (selection criterion: overall P<0.05) was used to determine the best fitting regression model.
The same multilevel regression procedure was conducted for the subgroup of observers on which we had data regarding the prevalence of hazardous attitudes (n = 62), adding the attitudes as an extra predictor. None of the hand surgeons had hazardous levels of impulsivity and/or resignation/external locus of control; therefore, these variables could not be entered into the bivariable multilevel analysis.
Characteristics of observers
A total of 92 of the 283 invited hand surgeons (33%) completed the study, 20 hand surgeons only partially completed the survey and were excluded from analysis, one opted out and 170 surgeons (60%) did not respond. The response rate should be viewed in light of the known high rate of unresponsiveness among the members of the SOVG; this is due to out of date email addresses and a subgroup of inactive members that rarely participate. Of the hand surgeons who completed the study, 42 observers were assigned to review clinical information alone and 50 to review clinical information as well as radiographs [Table 1].
Table 1.
Surgeon Characteristics (n = 92)
|
Total
(n = 92) |
Clinical information only
(n = 42) |
Clinical information and radiographs
(n = 50) |
||||
|---|---|---|---|---|---|---|
| Parameter | Number | % | Number | % | Number | % |
| Sex | ||||||
| Male | 79 | 85.9 | 37 | 88.1 | 42 | 84.0 |
| Female | 13 | 14.1 | 5 | 11.9 | 8 | 16.0 |
| Location of practice | ||||||
| North America | 73 | 79.3 | 33 | 78.6 | 40 | 80.0 |
| Europe and United Kingdom | 9 | 9.8 | 3 | 7.1 | 6 | 12.0 |
| Central and South America | 7 | 7.6 | 3 | 7.1 | 4 | 8.0 |
| Other | 3 | 3.3 | 3 | 7.1 | 0 | 0.0 |
| Training | ||||||
| Orthopaedic Surgery | 80 | 87.0 | 35 | 83.3 | 45 | 90.0 |
| Plastic Surgery | 10 | 10.9 | 5 | 11.9 | 5 | 10.0 |
| General Surgery | 2 | 2.2 | 2 | 4.8 | 0 | 0.0 |
| Specialization | ||||||
| Hand Surgery | 90 | 97.8 | 40 | 95.2 | 50 | 100.0 |
| General Orthopaedic Surgery | 2 | 2.2 | 2 | 4.8 | 0 | 0.0 |
| Years in independent practice | ||||||
| 1–5 | 19 | 20.7 | 4 | 9.5 | 15 | 30.0 |
| 6–10 | 21 | 22.8 | 12 | 28.6 | 9 | 18.0 |
| 11–20 | 34 | 37.0 | 17 | 40.5 | 17 | 34.0 |
| >20 | 18 | 19.6 | 9 | 21.4 | 9 | 18.0 |
| Practice setting | ||||||
| Hospital-employed, academic | 53 | 57.6 | 22 | 52.4 | 31 | 62.0 |
| Hospital-employed, non-academic | 10 | 10.9 | 5 | 11.9 | 5 | 10.0 |
| Private practice | 29 | 31.5 | 15 | 35.7 | 14 | 28.0 |
| Supervising surgical trainees in operating room | ||||||
| No | 16 | 17.4 | 8 | 19.0 | 8 | 16.0 |
| Yes | 76 | 82.6 | 34 | 81.0 | 42 | 84.0 |
| Number of patients typically treated annually with TMC arthrosis | ||||||
| 1–20 | 8 | 8.7 | 4 | 9.5 | 4 | 8.0 |
| 21–50 | 24 | 26.1 | 9 | 21.4 | 15 | 30.0 |
| >50 | 60 | 65.2 | 29 | 69.0 | 31 | 62.0 |
| Hazardous attitudes (n = 62) | ||||||
| Macho | ||||||
| No | 45 | 72.6 | 25 | 75.8 | 20 | 69.0 |
| Yes | 17 | 27.4 | 8 | 24.2 | 9 | 31.0 |
| Self-confidence | ||||||
| No | 57 | 91.9 | 28 | 84.8 | 29 | 100.0 |
| Yes | 5 | 8.1 | 5 | 15.2 | 0 | 0.0 |
| Worry/anxiety | ||||||
| No | 58 | 93.5 | 30 | 90.9 | 28 | 96.6 |
| Yes | 4 | 6.5 | 3 | 9.1 | 1 | 3.4 |
| Antiauthority | ||||||
| No | 59 | 95.2 | 31 | 93.9 | 28 | 96.6 |
| Yes | 3 | 4.8 | 2 | 6.1 | 1 | 3.4 |
| Impulsive | ||||||
| No | 62 | 100.0 | 33 | 100.0 | 29 | 100.0 |
| Yes | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 |
| Resignation/external locus of control | ||||||
| No | 62 | 100.0 | 33 | 100.0 | 29 | 100.0 |
| Yes | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 |
Results
Thirteen variables met the criterion of P<0.15 in bivariable analysis and were entered into the backwards stepwise multivariable regression analysis [Table 2]. Accounting for potential interaction among variables using multilevel regression, the factors independently associated with a higher likelihood of a recommendation for surgery included: pain with a greater limiting effect on daily activities, previous diagnosis of TMC arthrosis (1 year ago compared to no previous diagnosis), previous treatment of TMC arthrosis (previous splint or splint and cortisone injection compared to no previous treatment), greater pain with the grind test, adduction contracture and metacarpophalangeal hyperextension, viewing radiographs of the thumb, location of surgeon’s practice (Europe and United Kingdom or Central and South America compared to North America), and surgeon training (Plastic Surgery compared to Orthopaedic Surgery or Plastic Surgery compared to General Surgery) [Table 3].
Table 2.
Bivariable Regression Analyses for Operative Treatment of TMC Arthrosis (92 Observers)
| Patient parameter | F | df | P value |
|---|---|---|---|
| Sex | 3.6 | 1 | 0.058 |
| Age | 1.8 | 1 | 0.18 |
| Dominant thumb | 2.8 | 1 | 0.092 |
| Daily activities | 37 | 4 | < 0.001 |
| Diagnosis of TMC arthrosis | 2.5 | 2 | 0.082 |
| Previous treatment for TMC arthrosis | 270 | 2 | < 0.001 |
| Shoulder sign | 0.45 | 1 | 0.50 |
| Pain with the grind test | 4.9 | 4 | < 0.001 |
| Crepitation with the grind test | 0.26 | 4 | 0.91 |
| Adduction contracture and MCP hyperextension | 5.7 | 1 | 0.017 |
| Pinch and grip strength | 1.2 | 1 | 0.27 |
| Radiographic parameter | |||
| Radiographs (yes / no) | 5.5 | 1 | 0.019 |
| Surgeon parameter | |||
| Sex | 3.5 | 1 | 0.062 |
| Location of practice | 9.7 | 3 | < 0.001 |
| Training | 6.3 | 2 | 0.0019 |
| Specialization | 0.17 | 1 | 0.68 |
| Years in independent practice | 0.45 | 3 | 0.72 |
| Practice setting | 3.0 | 2 | 0.048 |
| Supervising surgical trainees in operating room | 1.7 | 1 | 0.19 |
| Number of patients typically treated annually with TMC arthrosis | 6.7 | 2 | 0.0013 |
DF = degrees of freedom. Numbers in bold indicate significance at P < 0.05.
Table 3.
Best Multilevel Regression Model for Operative Treatment of TMC Arthrosis (92 Observers)
| Patient parameter | Rate of surgery recommendation | F | P value | Odds ratio | 95% confidence interval | |
|---|---|---|---|---|---|---|
| Lower | Upper | |||||
| Daily activities | 46 | < 0.001 | ||||
| The pain does not limit daily activities | 22% | Reference category | Reference category | Reference category | Reference category | |
| The pain slightly limits daily activities | 31% | 0.0017 | 1.8 | 1.2 | 2.6 | |
| The pain moderately limits daily activities | 39% | < 0.001 | 4.0 | 2.8 | 5.8 | |
| The pain severely limits daily activities | 48% | < 0.001 | 7.3 | 5.1 | 11 | |
| The pain makes it impossible to perform activities of daily living | 50% | < 0.001 | 8.5 | 5.9 | 12 | |
| Diagnosis of TMC arthrosis | 3.6 | 0.027 | ||||
| No previous diagnosis | 35% | Reference category | Reference category | Reference category | Reference category | |
| 3 months ago | 38% | 0.24 | 1.2 | 0.90 | 1.5 | |
| 1 year ago | 40% | 0.0074 | 1.4 | 1.1 | 1.9 | |
| Previous treatment for TMC arthrosis | 271 | < 0.001 | ||||
| No previous treatment | 15% | Reference category | Reference category | Reference category | Reference category | |
| Splint | 31% | < 0.001 | 4.0 | 2.9 | 5.4 | |
| Splint and cortisone injection | 67% | < 0.001 | 44 | 32 | 62 | |
| Pain with grind test | 6.3 | < 0.001 | ||||
| No pain | 32% | Reference category | Reference category | Reference category | Reference category | |
| Mild pain | 35% | 0.020 | 1.5 | 1.1 | 2.2 | |
| Moderate pain | 41% | 0.0012 | 1.8 | 1.3 | 2.5 | |
| Severe pain | 40% | < 0.001 | 2.0 | 1.4 | 2.9 | |
| Extreme pain | 41% | < 0.001 | 2.2 | 1.6 | 3.2 | |
| Adduction contracture and MCP hyperextension | 6.8 | 0.0090 | ||||
| No | 36% | Reference category | Reference category | Reference category | Reference category | |
| Yes | 40% | 0.0090 | 1.3 | 1.1 | 1.7 | |
| Radiographic parameter | ||||||
| Radiographs | 6.7 | 0.010 | ||||
| No | 32% | Reference category | Reference category | Reference category | Reference category | |
| Yes | 42% | 0.010 | 2.3 | 1.2 | 4.2 | |
| Surgeon parameter | ||||||
| Location of practice | 6.3 | < 0.001 | ||||
| North America | 32% | Reference category | Reference category | Reference category | Reference category | |
| Europe and United Kingdom | 65% | 0.0025 | 5.3 | 1.8 | 15 | |
| Central and South America | 60% | < 0.001 | 6.8 | 2.2 | 21 | |
| Other | 36% | 0.53 | 1.7 | 0.31 | 9.8 | |
| Training | 4.2 | 0.014 | ||||
| Orthopaedic Surgery | 36% | 0.20 | 4.4 | 0.46 | 42 | |
| Plastic Surgery | 57% | 0.023 | 17 | 1.5 | 196 | |
| General Surgery | 18% | Reference category | Reference category | Reference category | Reference category | |
Numbers in bold indicate significance at P < 0.05.
On average, surgeons preferred operative treatment for 38% of patients, but the rate of surgical treatment recommendation varied greatly; there was one surgeon who preferred operative treatment for all patients (100%), and two surgeons who preferred nonoperative treatment for all patients (0%).
Discussion
There is substantial and unexplained variation in rates of surgery for TMC arthrosis from surgeon to surgeon. Such variation generally indicates that the preferences and values of the surgeon have a greater influence on treatment decisions than the preferences and values of the patient (7). The latter was not investigated in this study. The role of surgery in patients with TMC arthrosis and the optimal technique are open to debate (5, 8). We studied the factors that have the greatest influence on surgeon recommendations to better aid efforts to reduce surgeon-to-surgeon variation in treatment of TMC arthrosis.
This study has several limitations. First, only surgeons participating in the SOVG were evaluated and a substantial portion of the SOVG members rarely or never participates. Secondly, due to the randomization of fictive patient characteristics (except for sex and age) there might have been one or more combinations of clinical information that would occur infrequently. Third, only a small amount of clinical information was presented. Occupation, avocation, and other information might influence the decision-making process. Fourth, observers responded to the question about their preferred treatment for a particular patient by picking their preferred treatment(s) from a list or by choosing “other” and describing the treatment. When a surgeon chose any kind of surgical option, we considered it “operative” treatment unless they mentioned that there was no absolute indication or urgency for surgery. Finally, we did not account for patient’s preferences and wishes.
In this study, surgeons were more likely to recommend surgery for patients with greater symptoms and disability, a prior diagnosis or treatment, and when presented with radiographs. The influence of radiographic appearance represents an unhelpful cognitive bias. There is mounting evidence that symptom intensity, magnitude of disability, and dissatisfaction with treatment are related to symptoms of depression and ineffective coping strategies, with relatively limited relation to objective impairment and pathophysiology (8-10). Surgeons have long recognized that symptoms and disability do not correlate with radiographic severity of arthrosis (11-13). We see patients every day that have severe TMC arthrosis they are not even aware of, and we have found that variation corresponds with effectiveness of coping strategies (14). There is sufficient evidence that we hope surgeons will reconsider and be more cautious with recommendations based on the misconception that greater illness always reflects greater disease and that treatment of the pathophysiology will always relieve symptoms and limit disability. Only a small degree of regional variation in surgery rates for common conditions can be explained by differences in illness burden, diagnostic practices, and patient attitudes about medical intervention (7). Studies have shown that surgeons’ beliefs, preferences and attitudes, and the extent to which patient preferences are incorporated into the decision-making process, better explain treatment variation (7, 15-17). Even though our results did not show a relationship between hazardous attitudes and treatment preference, location of practice and surgical training might partially reflect surgeons’ beliefs, preferences and attitudes. A survey among orthopaedic surgeons and referring physicians (rheumatologists and family physicians) for total knee arthroplasty found that (1) the opinion of the referring physicians varied greater than those of the surgeons; (2) the reasons for referral by referring physicians differed from the reasons the surgeons opted for total knee arthroplasty; and (3) half of the total variability of the most appropriate candidate (based on patient characteristics) for total knee arthroplasty was due to inconsistent individual physicians’ opinions (18). The latter is something that could potentially also play a role in the treatment decision-making for TMC arthrosis.
Our findings are consistent with studies that found substantial variation and inconsistency of recommendations for adenotonsillectomy (19) and knee arthroplasty (18, 20). Radiographic severity of knee arthrosis influences orthopaedic surgeon and even more so rheumatologist recommendations for total knee arthroplasty (18, 21). A retrospective study by Ochtman et al. (22) found substantial surgeon-to-surgeon variation in treatment strategies and rates of surgery for TMC arthrosis. A larger retrospective study by Becker et al. (2) showed that a younger patient age, seeing a second surgeon, and treatment by specific hand surgeons were associated with a greater likelihood of undergoing surgery for hand arthrosis. In a survey of 1,156 active members of the American Society for Surgery of the Hand nearly half (46%) chose to continue nonoperative treatment for a 42-year-old woman with moderate to severe pain and minimal radiographic changes, who was also unsatisfied with splinting and one corticosteroid injection (3). These hand surgeons were more likely from the mid-Atlantic compared to all other regions (60% and 45%, respectively) and more likely in practice for ≤15 years compared to those in practice for 16 to 25 years (52% and 46%, respectively) (3).
Surgeon-to-surgeon variation in recommendations for surgery relate most strongly to each surgeon’s circumstances and training as well as the influence of greater symptoms and disability, and exposure to radiographs. Given these notable surgeon biases, methods for increasing patient participation in the decision-making process merit additional attention and study. So-called decision aids are videos, websites, pamphlets, etc. that intend to provide patients with balanced, dispassionate, understandable information about their problem and the treatment options. The use of decision aids has resulted in a reduction of discretionary, preference-sensitive surgery in other fields (23), presumably at least in part because patients might recognize their maladaptive thoughts and emotions and be open to modifying them before undergoing the risks, discomforts, and inconveniences of surgery. In other words, well informed patients might be less likely to place all of their hope in invasive, passive treatments and more likely to feel hopeful no matter what they choose.
Table 4.
Bivariable Regression Analyses for Operative Treatment of TMC Arthrosis Including Hazardous Attitudes (62 Observers)
| Surgeon parameter | F | df | P value |
|---|---|---|---|
| Hazardous attitudes | |||
| Macho | 1.1 | 1 | 0.30 |
| Self-confidence | 1.8 | 1 | 0.18 |
| Worry/anxiety | 0.63 | 1 | 0.43 |
| Antiauthority | 1.6 | 1 | 0.21 |
DF = degrees of freedom.
Acknowledgements
We thank the following members of the Science of Variation Group for participating in this study: Mahmoud I. Abdel-Ghany; Joshua M. Abzug; Julie Adams; Ngozi M. Akabudike; Thomas Apard; L.C. Bainbridge; H. Brent Bamberger; Mark Baratz; Camilo Jose Romero Barreto; Taizoon Baxamusa; Ramon de Bedout; Steven Beldner; Prosper Benhaim; Philip Blazar; Martin Boyer; Maurizio Calcagni; Ryan P. Calfee; John T. Capo; Charles Cassidy; Louis Catalano III; Karel Chivers; Gregory DeSilva; Seth Dodds; David M. Edelstein; John M. Erickson; Peter J. Evans; Carlos H. Fernandes; R. Glenn Gaston; Richard S. Gilbert; Michael W. Grafe; Robert R.L. Gray; H.W. Grunwald; Andrew P. Gutow; Peter Hahn; Warren C. Hammert; Randy Hauck; Stuart M. Hilliard; Eric Hofmeister; Jerry I. Huang; Richard L. Hutchison; Asif Ilyas; Sidney M. Jacoby; Peter Jebson; Christopher M. Jones; David M. Kalainov; F. Thomas D. Kaplan; Saul Kaplan; Stephen A. Kennedy; Michael W. Kessler; Ryan Klinefelter; Jason H. Ko; Gerald A. Kraan; Steve Kronlage; Amy Ladd; Lewis B. Lane; Kendrick Lee; Paul A. Martineau; John McAuliffe; Greg Merrell; L.P. van Minnen; Cesar Dario Oliveira Miranda; Constanza L. Moreno-Serrano; Michael Nancollas; Luis Felipe Naquira Escobar; Daniel A. Osei; Patrick W. Owens; Bradley A. Palmer; M. Jason Palmer; Daniel Polatsch; Marco Rizzo; Craig Rodner; Tamara D. Rozental; David Ruchelsman; Kevin M. Rumball; Oleg M. Semenkin; Russell Shatford; Todd Siff; Robert R. Slater, Jr; Maximillian Soong; Sander Spruijt; Fabio Suarez; Carrie Swigart; John Taras; Andrew L. Terrono; Thomas F. Varecka; Erik T. Walbeehm; Frank L. Walter; Lawrence Weiss; Brian P.D. Wills; Jeffrey Wint; Jennifer Moriatis Wolf; Theresa Wyrick.
SJEB is supported by Dutch research grants from Anna Foundation|NOREF, Genootschap Noorthey, Stichting Fonds Doctor Catharine van Tussenbroek, and Stichting Vreedefonds, the Netherlands, for scientific research.
We thank Nan van Geloven, PhD, biostatistician at the Clinical Research Unit of the Academic Medical Center in Amsterdam for her help with the statistical analysis of this study.
Patient Consent:
Informed consent was waived for this study.
Disclosure:
The authors declare that they have no conflict of interest.
References
- 1.Spaans AJ, van Laarhoven CM, Schuurman AH, van Minnen LP. Interobserver agreement of the Eaton-Littler classification system and treatment strategy of thumb carpometacarpal joint osteoarthritis. The Journal of hand surgery. 2011;36(9):1467–70. doi: 10.1016/j.jhsa.2011.06.017. [DOI] [PubMed] [Google Scholar]
- 2.Becker SJ, Teunis T, Blauth J, Kortlever JT, Dyer GS, Ring D. Medical services and associated costs vary widely among surgeons treating patients with hand osteoarthritis. Clinical Orthopaedics and Related Research®. 2015;473(3):1111–7. doi: 10.1007/s11999-014-3912-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wolf JM, Delaronde S. Current trends in nonoperative and operative treatment of trapeziometacarpal osteoarthritis: a survey of US hand surgeons. The Journal of hand surgery. 2012;37(1):77–82. doi: 10.1016/j.jhsa.2011.10.010. [DOI] [PubMed] [Google Scholar]
- 4.Vermeulen GM, Slijper H, Feitz R, Hovius SE, Moojen TM, Selles RW. Surgical management of primary thumb carpometacarpal osteoarthritis: a systematic review. The Journal of hand surgery. 2011;36(1):157–69. doi: 10.1016/j.jhsa.2010.10.028. [DOI] [PubMed] [Google Scholar]
- 5.Wajon A, Vinycomb T, Carr E, Edmunds I, Ada L. Surgery for thumb (trapeziometacarpal joint) osteoarthritis. Cochrane database of systematic reviews. 2015(2) doi: 10.1002/14651858.CD004631.pub4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Guitton TG, Ring D, Earp BE, Ladd AL, Evans PJ, Kuo CE, et al. Interobserver reliability of radial head fracture classification: two-dimensional compared with three-dimensional CT. JBJS. 2011;93(21):2015–21. doi: 10.2106/JBJS.J.00711. [DOI] [PubMed] [Google Scholar]
- 7.Birkmeyer JD, Reames BN, McCulloch P, Carr AJ, Campbell WB, Wennberg JE. Understanding of regional variation in the use of surgery. The Lancet. 2013;382(9898):1121–9. doi: 10.1016/S0140-6736(13)61215-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Becker SJ, Bot AG, Curley SE, Jupiter JB, Ring D. A prospective randomized comparison of neoprene vs thermoplast hand-based thumb spica splinting for trapeziometacarpal arthrosis. Osteoarthritis and Cartilage. 2013;21(5):668–75. doi: 10.1016/j.joca.2013.02.006. [DOI] [PubMed] [Google Scholar]
- 9.Lozano-Calderon SA, Souer JS, Jupiter JB, Ring D. Psychological differences between patients that elect operative or nonoperative treatment for trapeziometacarpal joint arthrosis. Hand. 2008;3(3):271–5. doi: 10.1007/s11552-008-9098-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Makarawung DJ, Becker SJ, Bekkers S, Ring D. Disability and pain after cortisone versus placebo injection for trapeziometacarpal arthrosis and de Quervain syndrome. Hand. 2013;8(4):375–81. doi: 10.1007/s11552-013-9529-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Eaton RG, Glickel SZ. Trapeziometacarpal osteoar-thritis Staging as a rationale for treatment. Hand clinics. 1987;3(4):455–71. [PubMed] [Google Scholar]
- 12.Glickel SZ, Kornstein AN, Eaton RG. Long-term follow-up of trapeziometacarpal arthroplasty with coexisting scaphotrapezial disease. The Journal of hand surgery. 1992;17(4):612–20. doi: 10.1016/0363-5023(92)90303-7. [DOI] [PubMed] [Google Scholar]
- 13.Haara MM, Heliövaara M, Kröger H, Arokoski JP, Manninen P, Kärkkäinen A, et al. Osteoarthritis in the carpometacarpal joint of the thumb: prevalence and associations with disability and mortality. JBJS. 2004;86(7):1452–7. doi: 10.2106/00004623-200407000-00013. [DOI] [PubMed] [Google Scholar]
- 14.Becker SJ, Makarawung DJ, Spit SA, King JD, Ring D. Disability in patients with trapeziometacarpal joint arthrosis: incidental versus presenting diagnosis. The Journal of hand surgery. 2014;39(10):2009–15. doi: 10.1016/j.jhsa.2014.07.009. [DOI] [PubMed] [Google Scholar]
- 15.Janssen SJ, Teunis T, Guitton TG, Ring D. Do surgeons treat their patients like they would treat themselves? Clinical Orthopaedics and Related Research®. 2015;473(11):3564–72. doi: 10.1007/s11999-015-4304-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Teunis T, Janssen S, Guitton TG, Ring D, Parisien R. Do orthopaedic surgeons acknowledge uncertainty? Clinical Orthopaedics and Related Research®. 2016;474(6):1360–9. doi: 10.1007/s11999-015-4623-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Teunis T, Janssen SJ, Guitton TG, Vranceanu AM, Goos B, Ring D. Surgeon personality is associated with recommendation for operative treatment. Hand. 2015;10(4):779–84. doi: 10.1007/s11552-015-9755-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Toronto Arthroplasty Research Group Writing Committee. Variability in physician opinions about the indications for knee arthroplasty. The Journal of arthroplasty. 2011;26(4):569–75. doi: 10.1016/j.arth.2010.04.028. [DOI] [PubMed] [Google Scholar]
- 19.Bloor MJ, Venters GA, Samphier ML. Geographical variation in the incidence of operations on the tonsils and adenoids: An epidemiological and sociological investigation (Part 2) The Journal of Laryngology & Otology. 1978;92(10):883–95. doi: 10.1017/s0022215100086266. [DOI] [PubMed] [Google Scholar]
- 20.Wright JG, Coyte P, Hawker G, Bombardier C, Cooke D, Heck D, et al. Variation in orthopedic surgeons’ perceptions of the indications for and outcomes of knee replacement. CMAJ: Canadian Medical Association Journal. 1995;152(5):687. [PMC free article] [PubMed] [Google Scholar]
- 21.Fraenkel L, Suter L, Weis L, Hawker GA. Variability in recommendations for total knee arthroplasty among rheumatologists and orthopedic surgeons. The Journal of rheumatology. 2014;41(1):47–52. doi: 10.3899/jrheum.130762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ochtman AE, Guitton TG, Buijze GA, Zurakowski D, Mudgal C, Jupiter JB, et al. Trapeziometacarpal arthrosis: predictors of a second visit and surgery. Journal of hand and microsurgery. 2013;5(1):9–13. doi: 10.1007/s12593-012-0087-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Stacey D, Legare F, Col NF, Bennett CL, Barry MJ, Eden KB, et al. Decision aids for people facing health treatment or screening decisions. The Cochrane database of systematic reviews. 2014;1:CD001431. doi: 10.1002/14651858.CD001431.pub4. [DOI] [PubMed] [Google Scholar]