Keywords
pain - tendon lesions - VISA-P - pain pressure threshold - patellar tendinopathy
Introduction
The presence of disorders in the relationship between the structure and physiology
of the tendon can lead to varied clinical presentations for tendinopathies, characterized
by pain and tendon dysfunction. Moreover, it is considered that the increase of different
sensitizing substances in the tendon may significantly contribute towards tendinopathy.[1] As such, this increase of sensitizing substances is directly related to the response
to certain exploratory tests, such as pain in response to pressure or a decreased
pain pressure threshold.
Clinically, patellar tendinopathy (PT) often presents as a localized pain in the proximal
insertion of the patellar tendon on the patella. Despite the fact that its prevalence
is difficult to determine (as those who suffer from PT can usually continue training
and competing[2]), a study performed with young amateur athletes demonstrated that the prevalence
of PT varies across different sports, showing that basketball players have a prevalence
of 31.9%.[3]
[4]
[5]
Among professional basketball players, pain at the level of the patellar tendon can
lead to limitations in performance during training, which means that many players
have to decrease their level of training and competition during extended periods of
time, leading to a loss of physical capacity.[4] Physiotherapy interventions currently include prevention measures, as well as an
early diagnosis and treatment of PT.[6] Palpation tenderness is one of the most common signs used for the diagnosis of PT,
although it does not offer clinicians diagnostic certainty, as opposed to ultrasound
examinations.[7] However, recent studies show[8] that a pressure of 3.75 kg can be used as a discriminatory test between non-professional
players who are healthy and those with PT.
Most studies on the validity and reliability of pain pressure thresholds (PPTs) have
been performed with young amateur players and, on occasion, athletes;[5]
[8]
[9] however, it is unknown whether these data can be extrapolated to professional basketball
players to determine the presence of PT or to what extent PPTs may be useful for the
diagnosis of this population. In addition, despite the wide use of PPTs for assessment,
it is unknown whether these data can be extrapolated to professional basketball players
to determine the presence of PT or to what extent PPTs can be useful as a diagnostic
measure in this population. Furthermore, despite the extensive use of PPT for assessment
purposes, it is unknown whether other variables may influence the results of this
test and its accuracy to discriminate between knees with and without PT.
The main goal of this study was to identify whether the PPT can be used as a diagnostic
test of PT in professional basketball players, either on its own or combined with
other tests. Our hypothesis is that the PPT test has acceptable discriminatory accuracy
for the diagnosis of PT and that this accuracy improves if other significant covariates
are considered.
Material and Methods
Study Design and Participants
An observational study took place among professional basketball players in the first
division of the Spanish league. The study was presented to the Ethics Committee for
Human Studies of the Universitat de València and was approved with approval number
H1456389710571. The study was performed during the 2015/2016 basketball season, between
November and May. All teams belonging to the Association of Basketball Clubs (ACB,
in the Spanish acronym)—Spanish professional league—were verbally informed of the
study and invited to participate. The players were recruited by the physiotherapists
on their team with both the knowledge and consent of their trainer. After discussing
the study with the respective medical teams and team trainers, 8 teams, with a total
of 16 players, agreed to participate. The inclusion criteria consisted of: being a
professional player over the age of 18 and being an active player at the time of the
study. The exclusion criteria included players with a systemic illness; those who
had undergone knee surgery in the past; those who had received invasive physiotherapy
during the previous 3 months or those who had taken antiinflammatory medication during
the previous 24 hours. Also, players who were unable to communicate in Spanish or
English were excluded. All the selected players were informed verbally and in writing
of the study proposal and invited to participate voluntarily. All participants signed
an informed consent.
The players were diagnosed with PT using the following diagnostic criteria[8]: 1) previous history of pain in the patellar tendon during sports activities or
after the same, at least during the previous 3 months; 2) tenderness to pressure applied
to the patellar tendon at the inferior pole of the patella; and a VISA-P score of
80 or below.
Study Variables
The data collection was performed by two experienced physical therapists between 12
and 24 hours before the regular league game.
Pain Pressure Threshold
The PPT of the patellar tendons of the players was measured using an algometer (Pain
Test-Model FPK 40; Wagner Instruments, Greenwich, CT). The device was applied in the
tendon via its rubber disk with a surface of 1 cm2. All the algometry tests were performed by the same examiner (PMR), who was blinded
regarding the VISA-P scores and the visual analogue scale (VAS) results, which were
taken by another examiner (JBM). The PPT scores were measured with the players lying
in a supine position with a slight knee flexion (30°) maintained with the support
of a cushion placed in the popliteal space. Once the examiner located the site where
the player referred greater sensitivity or pain to palpation, he fixated the patella
with one hand from the superior pole and applied pressure with the algometer, progressively
increasing the compression by 0.1 kg/s increments until the player began feeling pain,
at which point the subject lifted his hand, and the examiner stopped applying compression,
recording the amount of pressure applied at that time. Two measurements were performed
at each point, separated by 60”, after which the mean of the two measurements was
recorded.
VAS
All players were asked which was the maximum level of pain they had experienced during
basketball training the previous week, which was registered using an 11 point (0-10)
VAS. The players were taught the scale with two faces, a happy face on one side and
a sad face on the other end. The scale is calibrated on the reverse side with a 0
(happy face) and a 10 (sad face).
VISA-P
The Victorian Institute of Sport Assessment, Patellar tendon (VISA-P) questionnaire
can be used to quantify the symptoms, function and capacity to perform sports activity
in those who suffer PT.[10] The player must respond to a series of questions which assesses their ability to
perform certain activities. Each question is assessed from 0 to 10 points, whereby
10 is complete functionality and 0 is total incapacity. The total sum of the different
activities on the scale gives a final score of the player's functional score, in which
100 is the maximum possible score. This questionnaire was not designed for the diagnosis
of PT; however, it is a useful tool to identify the clinical severity as well as the
symptoms, function and participation in sports.[3]
[4] For the Spanish players, we used the Spanish version of the VISA-P,[11] while for the remaining players, the English version was used.
Furthermore, 3 other covariates were measured:1) age (years); 2) height (cm); and
3) weight (kg).
Statistical Analysis
The t-test for independent samples was used, or the U Mann-Whitney test for examining the
differences between groups of tendons on the dominant side and the non-dominant side
in relation to the presence of absence of PT. Multiple linear regression models were
used to evaluate the covariates associated with the PPT scores of both the dominant
and non-dominant sides. Both models were produced by the retroactive method and used
the criteria of p < 0.10 for the exclusion of the model.
The receiver operating characteristic (ROC) curve was constructed with the data of
the area under the ROC curve (AUC) used to determine the precision of the PPT on its
own or combined with the remaining covariates. To determine the combined score, the
predicted probability of being able to discriminate between tendons with PT and healthy
tendons was calculated using logistic regression models (full model), in which the
dependent variable was PT (yes/no), and the independent variables were the PPT test
and the remaining covariates. Furthermore, other combined scores were considered using
only a partial number of covariates and the AUC was used to assess their discriminatory
presence. According to previous authors,[12] an AUC > 0.7 was used as a criterion for good discrimination.
Using the ROC analyses, we selected the best cut-off score for the PPT test when used
on its own and for each of the combined scores (the most efficient had an AUC > 0.8
using a smaller number of covariates). The best cut-off score was the value in which
sensitivity + specificity - 1 was maximized. Using these cut-off scores, the following
values were calculated: sensitivity (Se), specificity (Sp), positive likelihood ratio
(LR +) and negative likelihood ratio (LR-).
The calculation of the sample size was based on the general rule that 15 subjects
per predictor are necessary for a reliable equation in multivariate regression models.[13] A minimum of 75 participants were recruited assuming a maximum of five determinants.
All the analyses were performed using the SPSS version 19.0 software (IBM Corp., Chicago,
IL, USA).
Results
Of the 75 initial players in the study, 2 were excluded due to previous knee surgeries.
Finally, 73 players were included in the study, and 146 patellar tendons were analyzed.
The mean age of patients was 26.8 years (standard deviation [SD] 4.8); the mean height
was 198.1 cm (SD = 8.4); the mean weight was 95.5 kg (SD = 11.4) and the mean body
mass index was 24.3 kg/m2 (SD = 1.6). All athletes were able to both train and compete.
In total, 20 tendons (13.17% of the 146 students) were diagnosed with PT and the 126
remaining tendons were used as controls. No players presented PT in both knees at
the time of the study. Both the dominant and non-dominant knee were analyzed separately
for the descriptive analysis. In both cases, either on the dominant side or the non-dominant
side, the values of the VISA-P and VAS of the tendons with PT were significantly different
from those of the healthy tendons. However, the PPT values of the knees with PT were
only significantly lower than the healthy tendons on the non-dominant side ([Table 1]).
Table 1
|
|
PT
|
Healthy
|
Difference mean PT-healthy (95% CI)
|
|
Dominant leg
|
N
|
12
|
61
|
|
|
PPT (kg)
|
9.3 (3.1)
|
10.4 (2.5)
|
1.1 (-0.5 to 2.8)
|
|
VISA-P (points)
|
69.7 (8.7)
|
97.2 (4.7)
|
27.5 (21.8 to 33.1)*
|
|
VAS (cm)
|
3.7 (3.3)
|
0.4 (0.9)
|
-3.3 (-5.4 to -1.2)*
|
|
Non-dominant leg
|
N
|
8
|
65
|
|
|
PPT (kg)
|
7.2 (3.4)
|
10.7 (2.5)
|
3.5 (1.5 to 5.5)**
|
|
VISA-P (points)
|
73.0 (5.8)
|
97.2 (4.5)
|
24.2 (20.8 to 27.6)**
|
|
VAS (cm)
|
3.7 (2.1)
|
0.3 (0.9)
|
-3.7 (-5.1 to -1.6)**
|
Determinants of the PPT Values
The multivariate predictive models of the PPT values on both sides (dominant and non-dominant
side) were statistically significant (p < 0.05) accounting for 17.8% and a 23.1% of the variance (R2), respectively ([Table 2]). The height was positively associated both with the dominant and the non-dominant
sides, describing a similar percentage of the variance (partial R2), 8.8% and 7.2%,
respectively. The VAS score was negatively associated on both sides, accounting for
7.9% of the variance on the dominant side and 16% on the non-dominant side.
Table 2
|
Side
|
Independent variable
|
Estimate
|
SD
|
B standardized
|
P-value
|
R2 model
|
R2 partial
|
|
Dominant
|
17.8%
|
|
|
Height (cm)
|
0.09
|
0.04
|
0.29
|
0.008
|
|
8.8%
|
|
VAS score
|
−0.38
|
0.15
|
−0.28
|
0.012
|
|
7.9%
|
|
Non-dominant
|
23.1%
|
|
|
Height (cm)
|
0.09
|
0.04
|
0.27
|
0.013
|
|
7.2%
|
|
VAS score
|
−0.76
|
0.20
|
−0.40
|
0.000
|
|
16%
|
Discriminatory Accuracy of the Test on its Own or Combined
The ROC and AUC of the PPT test on its own and the combined values of the covariates
are shown in [Fig. 1]. The AUC for the PPT test was only 0.713 (95% CI 0.587–0.840), and when it was combined
with the significant covariates (height and VAS), this increased considerably until
0.867 (95% CI 0.759–0.975). The combined values of the PPT with the VAS alone also
offered a greater increase until 0.852 (95% CI 0.732–0.973). The AUC of the complete
combined values, without the VAS value, was 0.717, which means that the VAS test adds
a high value as a discriminatory marker of PT. Based on these results, we calculated
an optimal cut-off point for the test performed on its own or combined with the VAS
score. The cut-off point for a more precise discrimination of a tendon with PT was
8.8 kg for the PPT on its own and 14% for the PPT combined with the VAS. The use of
the equation
introducing the PPT and VAS values enables the possibility to obtain the probability
as a prognosis of the combined prognosis; therefore, if this is superior or equal
to 0.14, a PT may be diagnosed. The ratios of sensitivity, specificity and probability
calculated for these cut-off points are shown in [Table 3].
Fig. 1 Receiver-operating characteristic (ROC) curves, with the area under the ROC curve
(AUC), calculated for the pain pressure thresholds (PPTs) alone and with the predictive
probability combining the test and its covariates to determine patellar tendinopathy.
Abbreviations: VAS, visual analogue scale.
Table 3
|
Measures
|
PPT
|
|
PPT scores
|
Combined values PPT + VAS
|
|
Sensitivity
|
71.4
|
80
|
|
Specificity
|
55.5
|
90.5
|
|
LR+
|
1.59
|
8.42
|
|
LR-
|
0.52
|
0.219
|
Discussion
According to the information available, this is the first study to describe the PPT
value of professional basketball players with, and without, PT. Our results reveal
that the height and the values of the VAS are determining factors of the PPT values,
and that the PPT does not enable the possibility to precisely discriminate between
knees with, and without, PT on its own. However, the discriminative capacity and,
therefore, the diagnostic value of the PPT can be notably improved when it is combined
with the VAS covariate. In this study, we found that the PPT values did not differ
when comparing the dominant leg of players with PT with the values of the dominant
leg in players without PT. The height had the best predictive capacity on the dominant
side (8.8% and 7.2%), whereas the VAS value had a greater predictive value in the
non-dominant side (7.9% and 16%). Speculatively, these results can be related with
adaptations in the mechanisms of pain modulation on the dominant side, in which one
would expect more use of the same and, therefore, the existence of a greater number
of muscle contractions able to modulate pain.[14]
[15]
Clinically, the optimal cut-off point for the PPT necessary in order to consider the
diagnosis of PT was 8.8 kg and 14% for the combined scores using the PPT and the VAS.
Although these were the cut-off points for the diagnosis of PT, the performance of
a correct diagnosis necessarily requires this information to be combined with other
clinical evidence.
In our sample, we identified 20 tendons with PT and 126 healthy knees (20 athletes
with PT compared to 53 healthy athletes), representing a prevalence of 37.7% with
tendinopathy, slightly higher than the results obtained by Lian et al, who reported
a prevalence of 31.9% among elite players.[4] However, a study by Zwerver et al,[3] in non-elite athletes, found a prevalence of PT ranging between 2.5 and 14.4%, depending
on the sport they were performing, Cook et al found that, among adult basketball players,
this was 11.8%, whereas among junior basketball players under the age of 18, the prevalence
was 7%.[16] The differences in the results of our study can be explained by the characteristics
of the sample, as we included professional basketball players, who have significant
load requirements.
At present, it is not possible to affirm the existence of a direct relationship between
image assessment (ultrasound and magnetic resonance) and the symptoms of PT. Furthermore,
there is no relationship between the amount of pain and the degree of vascularization.[17] However, previous studies show that algometry is a valid and reliable tool in the
diagnosis of PT.[9] Besides, the PPT has proven to be the most reliable method for the assessment of
sensitivity to pain, compared with other methods, such as palpation or the VAS.[18] Notwithstanding, the cut-off point for the different parts of the body should be
calculated to enable us to perform a more specific assessment.[19] Likewise, each population should have their own cut-off point, as occurs with the
professional basketball players in our study. Sensitivity to pain is not a uniform
characteristic;[20] therefore, it is necessary to know the pain threshold of professional basketball
players in order to use this as a diagnostic test in this population. Regarding algometry,
Kregel et al[8] described a mean PPT of 2.2 kg/cm2 in the PT of university sportsmen, which is much lower than our result of 8.8 kg/cm2 found in the same tendon. We think that these data highlight the importance of using
specific values for professional basketball players.
The PPT test was unable to effectively discriminate the presence of PT in professional
basketball players because the values of LR+ (1.59) and LR- (0.52) were poor (LR+ < 2
and LR- > 0.50), despite having obtained an AUC higher than 0.7. The accuracy was
greater and positively improved with the inclusion of the VAS covariate. When this
occurs, the accuracy of a test depends on another covariate, and it is commonplace
to adjust to these variables in the statistical analysis.[21] Even so, according to our information, our study is the first to assess the influence
of covariates in the discriminatory accuracy of the PPT test. Although the adjustment
of the covariates is routine in other medical areas, this issue is not extensively
assessed as of yet in the functional tests used in sports medicine. Neither is it
common in therapeutic and etiological studies, as it has received little attention
in the development of markers for tests or diagnoses.[13] Our study has focused on this aspect, therefore contributing to redefining our understanding
regarding the way functional tests can be used in a combined manner in professional
players.
Study Limitations
The gold standard diagnosis of PT was based on a questionnaire (VISA-P) and tenderness
to palpation at the inferior pole of the patella performed by a physical therapist.
An image-based diagnosis was not performed, which, however, could have provided relevant
information for the definition of the clinical diagnosis and categorization of the
type of tendinopathy. However, many studies have included patients with PT[7]
[16] via a self-report questionnaire. Furthermore, although we have found determinants
of the PPT that improve the discrimination between knees with, or without, PT, it
is possible that other factors not considered in our study (for example, gender) may
increase the discriminative capacity of this test. Lastly, this study only included
players who were able to compete; therefore, the prevalence of PT may have been lower
than in other studies. Besides, the accuracy of the diagnosis could have improved
if the study had been performed on the whole team, without ruling out those who could
not compete.
Future studies could analyze the behavior of these same tests and the influence of
covariates in the female population, as our study was limited to a male population
with the aim of avoiding bias due to the existing differences between the genders
regarding the mechanisms of pain modulation. Furthermore, in this study, we researched
the pain that PT produces at a certain point in time; therefore, future research should
prospectively control and compare the levels of pain throughout the season.
Conclusions
The present study shows that the measurement of the PPT in the patellar tendon is
not recommended to be performed on its own for discriminating the presence of PT in
professional basketball players. However, when a cluster of the PPT and the VAS is
used for the diagnosis of PT in professional players, the diagnostic accuracy increases
considerably. Due to the good values of the LR+ obtained by the PPT and VAS cluster
for the diagnosis of PT, and the ease of its application, the combined use of these
variables is recommended for performing a clinical diagnosis of PT in professional
basketball players.
