Ulnar Collateral Ligament Reconstruction
in Major League Baseball Pitchers
Brett W. Gibson, MD, David Webner, MD, G. Russell Huffman, MD, MPH, and
Brian J. Sennett, MD
From the Department of Orthopaedic Surgery, Penn Sports Medicine Center,
Division of Sports Medicine, Philadelphia, Pennsylvania
Background: Ulnar collateral ligament reconstruction is commonly performed in major league pitchers, but little is known about
pitching performance after a return to major league play.
Hypothesis: Pitching performance after ulnar collateral ligament reconstruction returns to baseline by the second season after
surgery.
Study Design: Cohort study (prognosis); Level of evidence, 2.
Methods: Data were reviewed for 68 major league pitchers who pitched in at least 1 major league game before undergoing ulnar
collateral ligament reconstruction between 1998 and 2003. Mean innings pitched per season, earned run average, and walks and
hits per inning pitched were compared for each major league pitcher before and after surgery. All demographic and performance
variables were analyzed for an association with ulnar collateral ligament insufficiency and a successful return to major league play.
Results: Fifty-six (82%) pitchers returned to major league play at a mean of 18.5 months after surgery with no significant change
in mean earned run average or walks and hits per inning pitched. The mean innings pitched per season was not statistically different from controls by the second season after surgery. Starting pitchers demonstrated a higher risk of ulnar collateral ligament
injury requiring reconstruction. More experienced pitchers and those with a higher earned run average were less likely to require
ulnar collateral ligament reconstruction. No factors predictive of a successful return to play were identified.
Conclusion: Most major league pitchers return from ulnar collateral ligament reconstruction by the second season after surgery with
no statistical change in mean innings pitched, earned run average, or walks and hits per inning pitched from preinjury levels.
Keywords: elbow; ulnar collateral ligament; injury; pitcher; overuse
“Tommy John surgery.” Today, UCL reconstruction is commonly performed for overhead athletes with UCL injury.
Several published case series have described the results of
UCL reconstruction.1-3,7,13,15,18,19 Few have dealt exclusively
with major league pitchers, and little is known about potential risk factors for injury requiring reconstruction or factors
contributing to a successful return to play in this patient
population. Additionally, little is known about the performance of pitchers who return to major league play after UCL
reconstruction. The purpose of this study is to describe the
results of UCL reconstruction in major league pitchers, to
compare the performance of these pitchers before and after
surgery, and to identify risk factors for undergoing UCL
reconstruction and factors associated with a successful
return to major league play.
The primary stabilizer to valgus stress of the elbow is the
anterior oblique bundle of the ulnar collateral ligament
(UCL).9-12,14,16,17 Valgus stress on the medial elbow occurs
during the late cocking and acceleration phases of the overhead throwing motion. The anterior bundle of the UCL is
subject to high tensile stress during the acceleration phase of
throwing that may eventually lead to ligament attenuation
or failure.4-6,9-12,14,16,19 Ulnar collateral ligament insufficiency
is common among baseball pitchers and other overheadthrowing athletes. Historically, rupture of the UCL was
career-ending. In 1974, Jobe performed the first reconstruction of the anterior band of the UCL on Los Angeles Dodgers
pitcher Tommy John. Tommy John proceeded to win 164
games after the procedure, which has since been dubbed
*Address correspondence to Brett W. Gibson, MD, Department of
Orthopaedic Surgery, Penn Sports Medicine Center, Division of Sports
Medicine, 235 S. 33rd Street, Philadelphia, PA 19104-4405 (e-mail:
brett.gibson@uphs.upenn.edu).
No potential conflict of interest declared.
MATERIALS AND METHODS
Reconstructed Pitchers
Data were reviewed for 68 major league pitchers who
pitched in at least 1 major league game before undergoing
UCL reconstruction. No pitchers appeared in only 1 game
The American Journal of Sports Medicine, Vol. 35, No. 4
DOI: 10.1177/0363546506296737
© 2007 American Orthopaedic Society for Sports Medicine
575
576
Gibson et al
before reconstruction, and most pitchers appeared in several
games during multiple seasons. Reconstructed pitchers
were identified from team injury reports and press releases
indicating that the pitcher underwent UCL reconstruction.
“Tommy John surgery” was considered an acceptable reference to UCL reconstruction. Article XIII Section C of the
Major League Baseball Players Association Collective
Bargaining Agreement provides standards for injury reporting in major league baseball, stating that the “Application by
a Club to the Commissioner to place a Player on the
Disabled List shall be accompanied by a Standard Form of
Diagnosis” to be completed by the team physician. The
Standard Form of Diagnosis includes fields for UCL injury
and surgery in the arm/elbow diagnosis description. The
physician and a club official are required to sign the completed form before submission to the Commissioner. These
standards ensure a high degree of reliability for major
league baseball injury reports as a primary source of identifying the study population.
Surgery was performed between 1998 and 2003. Age and
major league pitching experience were determined at the
time of UCL reconstruction. Height, weight, throwing handedness, and pitching role (starting or relief pitcher) were
recorded for each reconstructed pitcher. Body mass index was
calculated from height and weight data. The date of return to
major league play was recorded, and the time to return was
calculated. The month of surgery was not available for 1 subject who successfully returned to major league play.
The innings pitched, earned run average (ERA), and walks
and hits per inning pitched (WHIP) during major league play
were recorded for each reconstructed pitcher during 7 consecutive seasons. Earned run average represents the average
number of earned runs allowed per 9 innings pitched, or 1
complete major league game. An earned run is any run
scored by the opposing team for which the pitcher is held
accountable. Runs scored as a result of fielding errors, including those by the pitcher himself, or passed balls are considered unearned and are excluded from the calculation. Walks
and hits per inning is determined by dividing the sum of
walks and hits by the total number of innings pitched. The
average during 3 consecutive seasons before the index year
was defined as the subject’s preindex performance. The average during 3 consecutive seasons after the index year was
defined as the pitcher’s postindex performance. Data from
the index season were excluded from both calculations.
Control Pitchers
A control group is necessary to define expected levels of
performance in major league pitchers during the course of
7 seasons. Every fifth name was selected from a complete
alphabetical roster of major league pitchers from the 2001
season for a total of 112 pitchers in the control group. The
2001 major league season was defined as the index year for
this cohort. Players with a known history of UCL reconstruction were excluded from this list before selection.
Pitchers were not excluded from the control group on the
basis of any other injuries or surgical procedures.
The American Journal of Sports Medicine
Age and major league pitching experience were determined at opening day in 2001. Height, weight, throwing
handedness, and pitching role (starting or relief pitcher)
were recorded for each control pitcher. Body mass index was
calculated from height and weight data. Preindex and
postindex performance averages were calculated as
described for reconstructed pitchers.
Statistical Methods
Paired analysis of preindex and postindex performance
measures was performed for each major league pitcher in
the reconstructed and control groups. Performance measures were compared between reconstructed and control
groups on a seasonal basis to better account for the expected
decline in innings pitched among reconstructed pitchers
during the recovery period. Age, experience, height, weight,
body mass index, throwing handedness, pitching role, and
preindex performance measures were then analyzed for an
association with risk of UCL injury requiring reconstruction
and a successful return to major league play.
The Student t test was used to compare age, experience,
height, weight, body mass index, and performance measures
between groups. Paired means testing was used to assess
preindex and postindex performance measures within each
group of pitchers. The Fisher’s exact probability test was
used to compare throwing handedness and pitching role
between groups. Both univariate and multivariate statistical analyses were performed using STATA 8.0 (StataCorp,
College Station, Texas).
RESULTS
Player Characteristics
Player characteristics are presented in Table 1. Sixty-eight
reconstructed pitchers and 112 control pitchers were
included in the study. The mean age was 28.2 years (range,
22-42 years) for reconstructed pitchers and 29.1 years
(range, 21-44 years) for controls (P = .26). The mean major
league experience was 5.1 years (range, 1-18 years) for
reconstructed pitchers and 5.9 years (range, 1-22 years)
for controls (P = .26). Thirty-nine reconstructed pitchers
(57%) were starting pitchers compared with 40 (36%) control pitchers (P = .005). By comparison, 232 major league
pitchers (38 percent) in 2005 were starting pitchers, which
is not statistically different from the control group (P = .67).
No statistical difference in height, weight, body mass index,
or throwing handedness was identified between the reconstructed and control groups (Table 1).
The preindex average innings pitched per season was
92.98 for reconstructed pitchers and 87.11 for controls (P =
.58). The preindex mean ERA was 4.16 for reconstructed
pitchers and 4.37 for controls (P = .03). The preindex mean
WHIP was 1.362 for reconstructed pitchers and 1.384 for
controls (P = .04).
The mean innings pitched, ERA, and WHIP for the control group was statistically representative of the mean for
Vol. 35, No. 4, 2007
UCL Reconstruction in Major League Baseball Pitchers
TABLE 1
Reconstructed and Control Pitcher Characteristics
Seasonal Comparison
140.00
Reconstructed
Control
UCL Reconstruction
(n = 68)
Control
(n = 112)
P Value
28.2 ± 0.4
5.1 ± 0.1
29.1 ± 0.3
5.9 ± 0.1
.26
.26
Age (years)
Major league
experience
(seasons)
Height (m)
Weight (kg)
Body mass
index
% Right-handed
pitchers
% Starting pitchers
Innings Pitched
120.00
Player
Characteristic
577
100.00
*
80.00
**
60.00
40.00
20.00
0.00
−3
1.89 ± 0.03
95.6 ± 1.4
26.9 ± 0.4
1.88 ± 0.02
94.7 ± 0.8
26.8 ± 0.2
.58
.54
.73
78%
74%
.60
57%
36%a
.005
a
38% of major league baseball pitchers in 2005 were starting
pitchers.
all major league pitchers who played during the 2001
major league baseball season. The overall league average
was calculated using the total innings pitched and the
total earned runs, walks, and hits allowed for all 622 major
league pitchers in 2001. The mean performance measures
for the control group during the index year were within
2.4% of the calculated league average.
Return to Major League Pitching
Fifty-six (82%) reconstructed pitchers returned to major
league play at a mean of 18.5 months (range, 10-49 months)
after surgery. Reconstructed pitchers were observed for an
average of 4.3 years (range, 2-8 years). No reconstructed
pitchers returned to major league play during the index
year. Thirty-two (47%) reconstructed pitchers returned to
major league play during the first postindex season. Fifty
(74%) reconstructed pitchers returned to major league play
by the second postindex season. Fifty-three (78%) reconstructed pitchers returned to major league play by the third
postindex season. The remaining 3 reconstructed pitchers
returned to major league play greater than 3 years after the
index year.
Player Performance
Fifty-four reconstructed pitchers and 70 control pitchers
were available for paired analysis of preindex and postindex performance measures. Preindex averages could not be
calculated for 1 reconstructed pitcher and 23 control pitchers who did not appear in any major league games before
the index year. Postindex averages could not be calculated
for 12 reconstructed pitchers who failed to return to major
league play after surgery, 3 reconstructed pitchers who
returned more than 3 seasons after the index year, and 13
control pitchers with no major league appearances during
the postindex period. Neither preindex nor postindex
−2
−1
0
Season
1
2
3
Figure 1. Comparison of innings pitched by individual season between reconstructed pitchers and controls. Season 0
represents the index year. *P = .06; **P < .05.
averages could be calculated for 6 control pitchers who
played during the index year only.
Paired analysis of reconstructed pitchers revealed that
innings pitched per season decreased significantly from
a preindex mean of 97.10 (range, 6.00-234.56) to 70.17
(range, 2.00-198.00) after the index year (P = .003). Control
pitchers also demonstrated a significant decrease in innings
pitched per season from a preindex mean of 94.73 (range,
1.00-246.44) to 79.29 (range, 4.00-238.22) after the index
year (P = .02). The ERA for reconstructed pitchers demonstrated no significant change from a preindex mean of 4.12
(range, 0.00-11.37) to 4.21 (range, 0.00-8.24) after the index
year (P = .14). Similarly, control pitchers demonstrated no
significant change in ERA from a preindex mean of 4.34
(range, 0.00-13.48) to 4.27 (range, 2.19-15.75) after the index
year (P = .47). The WHIP for reconstructed pitchers showed
no significant change from a preindex mean of 1.362 (range,
1.000-2.422) to 1.356 (range, 0.500-1.958) after the index
year (P = .83). Control pitchers also demonstrated no significant change in WHIP from a preindex mean of 1.376 (range,
0.922-2.622) to 1.349 (range, 0.795-2.750) after the index
year (P = .97).
The average innings pitched per season was compared
between reconstructed and control pitchers for each of
the 7 seasons evaluated in this study (Figure 1). Innings
pitched was not statistically different between reconstructed and control groups in any of the 3 preindex seasons (P = .58, P = .58, and P = .77, respectively). During the
index year, the mean innings pitched was 50.80 for reconstructed pitchers and 72.25 for controls (P = .06). Only 6
pitchers in the reconstructed group completed a full season
during the index year before undergoing surgery. During
the first postindex season, the mean innings pitched was
44.91 for reconstructed pitchers who had successfully
returned to major league play and 82.98 for controls (P =
.001). Of the 32 reconstructed pitchers who returned during the first postindex season, 13 (41%) missed at least half
of the season. In the second and third postindex seasons,
the mean innings pitched among reconstructed pitchers
was not statistically different from control pitchers (P = .71
and P = .76, respectively).
The mean ERA between cohorts was similarly compared
(Figure 2). In the first and second preindex seasons, no
578
Gibson et al
The American Journal of Sports Medicine
Seasonal Comparison
Seasonal Comparison
6.00
5.50
Reconstructed
Control
1.500
1.450
5.00
*
WHIP
ERA
1.550
Reconstructed
Control
4.50
*
1.400
4.00
1.350
3.50
1.300
1.250
3.00
−3
−2
−1
0
Season
1
2
-3
3
Figure 2. Comparison of ERA by individual season between
reconstructed pitchers and controls. Thirty-nine reconstructed
pitchers played during the index year (Season 0). *P < .05.
statistical difference was found in the mean ERA for reconstructed and control pitchers (P = .83 and P = .19, respectively). During the third preindex season (1 season before
the index year), the mean ERA for reconstructed pitchers
(4.19) was significantly better than the mean ERA for control pitchers (4.54, P = .01). During the index year, the mean
ERA was 5.20 for reconstructed pitchers and 4.45 for controls, but this difference was not statistically significant (P =
.92). No statistical difference was found in the mean ERA for
reconstructed and control pitchers in any of the 3 postindex
seasons (P = .31, P = .18, and P = .69, respectively).
The mean WHIP for reconstructed and control pitchers
is presented in Figure 3. In the first and second preindex
seasons, no statistical difference was found in the mean
WHIP for reconstructed and control pitchers (P = .89 and
P = .43, respectively). During the third preindex season (1
season before the index year), the mean WHIP for reconstructed pitchers (1.368) was significantly better than the
mean WHIP for control pitchers (1.417, P = .007). During
the index year, the mean WHIP for reconstructed pitchers
was 1.489 compared with 1.361 for controls, but this difference was not statistically significant (P = .48). No statistical difference was found in the mean WHIP for
reconstructed and control pitchers in any of the 3 postindex seasons (P = .11, P = .12, and P = .84, respectively).
Risk Factors for UCL Reconstruction
Univariate analysis (Table 2) demonstrated a significantly
increased risk of undergoing UCL reconstruction among
starting pitchers (odds ratio [OR] = 2.42, 95% confidence
interval [CI] 1.31-4.48, P = .005). A significant association
was also observed for preindex ERA (OR = 0.80 per earned
run, 95% CI 0.65-0.99, P = .04) and preindex WHIP (OR =
0.25 per unit, 95% CI 0.07-0.97, P = .05) with reduced odds
for undergoing UCL reconstruction.
Comparison of preindex performance measures between
starting and relief pitchers revealed that starters averaged
significantly more innings pitched per season (130.00 vs
59.00, P < .001) with a superior WHIP (1.391 vs 1.484, P =
.05). The mean ERA was not statistically different between
starters (4.44) and relievers (4.60, P = .63). Linear regression
-2
-1
0
Season
1
2
3
Figure 3. Comparison of WHIP by individual season between
reconstructed pitchers and controls. Thirty-nine reconstructed
pitchers played during the index year (Season 0). *P < .05.
TABLE 2
Univariate Analysis of Risk Factors for Pitchers
Undergoing UCL Reconstruction
Player
Characteristic
Age
Major league experience
Starting pitcher
Right-handed pitcher
Mean innings pitched
Mean earned run average
Mean walks and hits
per inning pitched
Odds Ratio
(95% CI)
P Value
0.96
0.96
2.42
1.23
1.00
0.80
0.25
0.90-1.03
0.90-1.03
1.31-4.48
0.60-2.52
0.97-1.01
0.65-0.99
0.07-0.97
.280
.259
.005
.562
.577
.037
.045
analysis demonstrated a significant association between
major league experience and preindex performance measures as well. The mean innings pitched was 4.79 innings
higher per year of major league experience (P < .001). The
mean WHIP was 0.012 units lower per year of major league
experience (P = .03). The mean ERA was 0.05 earned runs
lower per year of major league experience, but this was not a
statistically significant association (P = .13). Since these factors are potential confounding variables, a multivariate logistic regression analysis was performed.
Multivariate analysis adjusting for all demographic and
performance variables, including age, major league experience, BMI, throwing handedness, pitching role, and preindex performance measures, revealed that the only
significant risk factor for undergoing UCL reconstruction
was being a starting pitcher (OR = 2.62, 95% CI 1.03-6.67,
P = .04). In a backward stepwise multivariate logistic
regression analysis (Table 3) excluding variables with a
greater than 10% probability of an association by chance
alone (P > .10), only starting pitching was associated with an
increased likelihood of undergoing UCL reconstruction (OR
= 2.45, 95% CI 1.21-4.98, P = .01), while major league experience (OR = 0.88 per year, 95% CI 0.80-0.95, P = .002) and
a higher preindex ERA (OR = 0.76 per earned run, 95% CI
0.61-0.95, P = .02) were both associated with a significantly
decreased likelihood of undergoing the procedure.
Vol. 35, No. 4, 2007
UCL Reconstruction in Major League Baseball Pitchers
TABLE 3
Backward Multivariate Stepwise Logistic Regression
Analysis of Risk Factors for Pitchers Undergoing UCL
Reconstruction, Beginning With All Variables and
Eliminating Factors With P > .10 in a Stepwise Fashion
Player
Characteristic
Major league
experience (per year)
Starting pitcher
Mean earned run average
Odds Ratio
(95% CI)
P Value
0.88
0.80-0.95
.002
2.45
0.76
1.21-4.98
0.61-0.95
.013
.018
Factors Predicting Successful Return to MLB
Univariate analysis identified no significant factors predictive of a successful return to major league play among
reconstructed pitchers (Table 4). Similarly, multivariate
analysis adjusting for all demographic and performance
variables revealed no significant factors associated with a
successful return to major league play.
DISCUSSION
Jobe7 first reported his results for UCL reconstruction in
throwing athletes in 1986. Ten of 16 (63%) patients returned
to the same level of play with a minimum 2-year follow-up.
Conway et al3 reported on a series of 56 patients with 2- to
15-year (mean, 6.3 years) follow-up after UCL reconstruction.
Thirty-eight (68%) successfully returned to sports, including
12 of 16 (75%) major league baseball players with no prior
elbow procedures. Andrews and Timmerman19 reported that
12 of 14 (86%) athletes treated with UCL reconstruction
returned to play.
More recently, Azar et al2 reported on 59 throwing
athletes who underwent UCL reconstruction with 12- to
72-month (mean, 35.4 months) follow-up. Forty-eight (81%)
returned to play at an average of 9.8 months after surgery.
Thompson et al18 published a series on 33 athletes with
2- to 4-year follow-up. Twenty-seven (82%) were able to
return to their previous level of play.
Rohrbough et al15 reported that 33 of 36 (97%) patients
undergoing UCL reconstruction successfully returned to
play at 2- to 5-year (mean, 3.3 years) follow-up, including
all college or professional athletes. The study group
included 33 baseball players, 1 lacrosse player, 1 tennis
player, and 1 golfer. Twenty-seven baseball players were
pitchers, and 9 were professional baseball pitchers
(4 major league). In 2004, Petty et al13 reported the results
of UCL reconstruction in 27 high school baseball players,
24 of whom were pitchers. Twenty (74%) returned to play
at or above their preinjury level at an average of
11 months postoperatively.
No previously reported studies have dealt exclusively
with the results of UCL reconstruction in major league
pitchers, an elite group of throwing athletes. The rate of
return in this series (82%) was comparable to that in more
579
TABLE 4
Univariate Analysis of Factors Associated With Successful
Return to Major League Baseball Pitching After UCL
Reconstruction
Player
Characteristic
Age
Major league
experience
Relief pitcher
Right-handed pitcher
Mean innings pitched
Mean earned run average
Mean walks and hits
per inning pitched
Odds Ratio
(95% CI)
P value
0.99
1.04
0.86-1.15
0.89-1.22
.990
.602
1.61
1.22
1.00
1.05
4.76
0.43-5.98
0.28-5.23
0.99-1.02
0.71-1.56
0.16-144.60
.475
.787
.415
.812
.370
recently published series. The mean time to return in this
population (18.5 months) was longer than previously
reported averages,2,13 with 3 pitchers requiring more than
3 seasons to return. In cases of prolonged delays in return
to play, nonbiologic factors are presumed to play a role,
including substandard pitching performance in minor
league assignments and the major league roster’s limited
capacity to accommodate additional pitchers.
In all recently published series of UCL reconstruction,
success was determined by a return to the same level of play
for at least 1 year. This measure as originally described by
Conway et al3 is thought to control for subjective performance bias by considering only the achievable level and duration of competition. Innings pitched, ERA, and WHIP
represent additional objective measures for baseball pitchers that have not been previously considered.
In the present study, pitchers who successfully returned
to major league play after UCL reconstruction experienced
no significant decline from preindex performance. The average innings pitched among reconstructed pitchers reached
control levels by the second season after surgery. Both reconstructed and control pitchers experienced a significant
decline in innings pitched during the postindex period. The
decline in innings pitched among control pitchers may represent missed games due to other injuries or an expected
decline in games played as some pitchers retire. Other control pitchers might fail to meet the standards of major
league play and lose time to more promising prospects. Most
reconstructed pitchers who returned during the first postindex season failed to play a full season, resulting in statistically fewer innings pitched when compared with controls.
The mean ERA and WHIP among reconstructed pitchers,
both of which are weighted performance measures and
therefore may represent better indicators of performance
than innings pitched, were not statistically different from
controls by the first season after surgery.
Very little is known about risk factors for UCL insufficiency. Overuse was implicated as a primary risk factor for
development of UCL insufficiency in a recent series of high
school baseball pitchers.13 A significantly higher proportion of pitchers in this study’s UCL reconstruction cohort
580
Gibson et al
were starting pitchers compared with the nonreconstructed (control) cohort. Starting pitchers demonstrated a
significantly higher likelihood of undergoing UCL reconstruction on both univariate and multivariate analyses.
While limited conclusions about risk may be ascertained
from observational studies, this finding appears to support
the concept of overuse as a risk factor for injury. Starting
pitchers average more innings pitched per season and per
game than relief pitchers, placing greater demands on the
throwing elbow and making them more susceptible to UCL
injury. No significant association between preindex
innings pitched and UCL reconstruction was found on
either univariate or multivariate analysis.
On univariate analysis, a higher likelihood of undergoing
UCL reconstruction was found among pitchers with better
weighted performance measures, that is, lower ERA and
WHIP, during the preindex seasons. After controlling for
potential confounding factors, a better preindex ERA and
fewer years of major league experience were associated with
an increased risk of undergoing UCL reconstruction. The
style of pitching required to achieve a lower ERA, including
elevated pitch speed and increased number of breaking
pitches thrown, might explain the higher rate of UCL injury
in this population. More experienced pitchers might demonstrate superior pitching mechanics that are protective with
regard to injury, with this improved durability translating
into longer major league careers. Younger pitchers might
choose to ignore early signs of UCL insufficiency and play
through the pain, resulting in more severe injury over time.
Finally, older pitchers faced with UCL reconstruction might
elect to retire rather than endure the prolonged recovery
time inherent in this procedure.
Univariate and multivariate analyses of potential factors contributing to a successful return to major league
play yielded no statistically significant results. Since relatively few pitchers fail to return, large numbers of reconstructed pitchers are required to achieve statistical
significance. Relief pitchers might enjoy an advantage over
starting pitchers given the lower demands placed on the
throwing elbow in terms of innings pitched per game. This
series demonstrated a trend toward more successful return
to play among relief pitchers, including 5 starting pitchers
who returned to major league play as relief pitchers, but
this difference was not statistically significant. A more successful return to major league play might be related to
unobserved factors such as surgeon experience, the method
of ligament reconstruction employed, or transposition of
the ulnar nerve at the time of reconstruction.
Limitations of the current study include the potential for
bias, confounding, and chance inherent in all observational
studies. In this report, potential sources of bias include missing data (information bias) that is unobtainable from injury
reports. We may not, in fact, be capturing all major league
pitchers undergoing UCL reconstruction during the study
period. In doing so, we may not be able to correctly infer as
to the performance, rate of return to pitching, and potential
risk factors for undergoing UCL reconstruction. Anecdotally,
UCL reconstruction is perceived favorably among major
league baseball players, with some major league pitchers
The American Journal of Sports Medicine
claiming that ball velocity and movement of breaking
pitches actually improve from preoperative levels after surgery. For this reason, players and teams have little motivation to misreport UCL injury compared with, for example,
rotator cuff injuries requiring repair, which have demonstrated much poorer results in this population.8 Confounding, or associations between risk factors for an outcome and
between 2 or more risk factors, is also possible in this type
of analysis leading to incorrect assumptions regarding associations of risk when they do not exist. To control for confounding, one may perform an experimental clinical study
(randomized, prospective clinical trial), but this is not possible when assessing risk. Alternatively, one may minimize
confounding by accounting for potential associated factors
using multivariate analysis as performed in this study.
Chance occurs when the probability of an association occurs
randomly. In an observational study, one may control for
chance only by selecting an acceptably stringent level of significance. In the present study, the significance of starting
pitchers having an increased likelihood of undergoing UCL
reconstruction was .005, or 5 in 1000 by chance alone. After
controlling for other potentially confounding factors, the
chance of this association persisting by chance alone
remained low at 13 in 1000 (P = .013).
In conclusion, the present study suggests that most
major league pitchers return from UCL reconstruction by
the second season after surgery with no discernable change
in mean innings pitched, ERA, or WHIP from preinjury levels. Starting pitchers are more likely to undergo UCL reconstruction but return to major league play at a rate not
statistically different from that of relief pitchers. Pitchers
with a lower preindex ERA or fewer years of major league
experience also demonstrated an increased risk of UCL
injury requiring reconstruction. No factors predictive of a
successful return to play were identified in this study.
REFERENCES
1. Andrews JR, Timmerman LA. Outcome of elbow surgery in professional baseball players. Am J Sports Med. 1995;23:407-413.
2. Azar FM, Andrews JR, Wilk KE, Groh D. Operative treatment of ulnar
collateral ligament injuries of the elbow in athletes. Am J Sports Med.
2000;28:16-23.
3. Conway JE, Jobe FW, Glousman RE, Pink M. Medial instability of the
elbow in throwing athletes: treatment by repair or reconstruction of
the ulnar collateral ligament. J Bone Joint Surg Am. 1992;74:67-83.
4. Dillman CJ, Fleisig GS, Andrews JR. Biomechanics of pitching with
emphasis upon shoulder kinematics. J Orthop Sports Phys Ther.
1993;18:402-408.
5. Fleisig GS, Andrews JR, Dillman CJ, Escamilla RF. Kinetics of baseball
pitching with implications about injury mechanisms. Am J Sports Med.
1995;23:233-239.
6. Hotchkiss RN, Weiland AJ. Valgus stability of the elbow. J Orthop Res.
1987;5:372-377.
7. Jobe FW, Stark H, Lombardo SJ. Reconstruction of the ulnar collateral ligament in athletes. J Bone Joint Surg Am. 1986;68:1158-1163.
8. Mazoue CG, Andrews JR. Repair of full-thickness rotator cuff tears in
professional baseball players. Am J Sports Med. 2006;34:182-189.
9. Morrey BF. Applied anatomy and biomechanics of the elbow joint.
Instr Course Lect. 1986;35:59-68.
10. Morrey BF, An KN. Articular and ligamentous contributions to the stability of the elbow. Am J Sports Med. 1983;11:315-319.
Vol. 35, No. 4, 2007
UCL Reconstruction in Major League Baseball Pitchers
581
11. Morrey BF, An KN. Functional anatomy of the ligaments of the elbow.
Clin Orthop Relat Res. 1985;201:84-90.
12. Morrey BF, Tanaka S, An K. Valgus stability of the elbow. Clin Orthop
Relat Res. 1991;265:187-195.
13. Petty DH, Andrews JR, Fleisig GS, Cain EL. Ulnar collateral ligament
reconstruction in high school baseball players: clinical results and
injury risk factors. Am J Sports Med. 2004;32:1158-1164.
14. Regan WD, Korinek SL, Morrey BF, An KN. Biomechanical study of
ligaments around the elbow joint. Clin Orthop Relat Res. 1991;271:
170-179.
rd
15. Rohrbough JT, Altchek DW, Hyman J, Williams RJ 3 , Botts JD.
Medial collateral ligament reconstruction of the elbow using the docking technique. Am J Sports Med. 2002;30:541-548.
16. Schwab GH, Bennett JB, Woods GW, Tullos HS. Biomechanics of
elbow instability: the role of the medial collateral ligament. Clin Orthop
Relat Res. 1980;146:42-52.
17. Søjbjerg JO, Ovesen J, Nielsen S. Experimental elbow instablitity after
transection of the medial collateral ligament. Clin Orthop Relat Res.
1987;218:186-190.
18. Thompson WH, Jobe FW, Yocum LA, Pink MM. Ulnar collateral ligament reconstruction in throwing athletes: muscle-splitting approach
without transposition of the ulnar nerve. J Shoulder Elbow Surg.
2001;10:152-157.
19. Timmerman LA, Andrews JR. Histology and arthroscopic anatomy of
the ulnar collateral ligament of the elbow. Am J Sports Med. 1994;22:
667-673.