We prospectively studied the cases of 199 consecutive patients (206
shoulders) with a chronic full-thickness rotator cuff tear (defined as
symptoms for more than three months) who were evaluated prior to surgery by
one surgeon during a forty-eight-month period. The patients were routinely
scheduled for outcome evaluations at three months, six months, and one year
after surgery. One hundred and twenty-five patients (125 shoulders) returned
for the one-year follow-up evaluation and were included in this study. (We
only included patients who returned for the one-year visit and who had a
chronic tear.) The study proposal was approved by the hospital institutional
review board. All of the patients underwent surgical treatment of the rotator
cuff tear, and the full-thickness tear was confirmed at the time of surgery.
The indications for surgery included failure of nonoperative treatment that
consisted of a physical therapy rehabilitation exercise program and, in some
cases, a corticosteroid injection. Patients with glenohumeral arthritis and
adhesive capsulitis were excluded.
The mean age of the patients was fifty-six years (range, thirty-two to
eighty years). Seventy-two shoulders (58%) were in men and fifty-three
shoulders (42%) were in women. Thirty-nine shoulders (31%) were in patients
who had a pending Workers' Compensation claim. The mean duration of the
symptoms (and standard deviation) prior to the evaluation was 16.04 ±
25.91 months (range, three to 210 months). Seventy-nine shoulders (63%) had a
history of an injury. The mean tear size was 2.2 cm (range, 1 to 4 cm).
Twenty-six shoulders (21%) had an open repair, sixty-two (50%) had a mini-open
repair, and thirty-seven (30%) had a completely arthroscopic repair. The
patients had a mean of 1.91 ± 1.51 medical comorbidities (range, zero
to six).
The preoperative and one-year postoperative evaluations included a detailed
medical history, a physical examination, and completion of a questionnaire
with a series of sections eliciting information regarding medical
comorbidities, shoulder pain, functional abilities, and general health status.
Demographic information (e.g., age and gender) as well as information
regarding Workers' Compensation status, the number of any previous surgical
procedures, smoking, duration of symptoms, and patient expectations were also
obtained with the series of questionnaires. Tear size and details of the
surgical technique were obtained from operative reports.
The first section of the patient self-assessment questionnaire addressed
general medical problems. The patients were asked a series of binary
(yes-or-no) questions regarding the presence of comorbid medical conditions.
This list of questions is included in the MODEMS (Musculoskeletal Outcomes
Data Evaluation and Management
System)12. The
patient was asked: "Do you have the problem (yes or no)?"
regarding (1) heart disease, (2) high blood pressure, (3) lung disease, (4)
diabetes, (5) ulcer or stomach disease, (6) kidney disease, (7) liver disease,
(8) anemia or other blood disease, (9) cancer, (10) depression, (11)
osteoarthritis or degenerative arthritis, (12) back pain, (13) rheumatoid
arthritis, or (14) other medical problem (please specify). The final inventory
included the presence or absence of hypertension, coronary artery disease,
other heart problems (valvular or arrhythmia), respiratory problems, pulmonary
embolus, diabetes, cancer, depression/anxiety, cerebral vascular accident,
headaches/migraines, liver disease, low-back pain, and arthritis. Each
"yes" answer was counted as one comorbidity. The number of
comorbidities was totaled for each patient.
The second section of the questionnaire included the Disabilities of the
Arm, Shoulder and Hand (DASH)
Questionnaire13.
The DASH was designed to assess symptoms and functional status, with an
emphasis on physical function, of patients with upper-extremity
musculoskeletal conditions. The DASH specifically questions patients regarding
pain, weakness, stiffness, ability to perform activities of daily living
(e.g., dressing, eating, and sleeping), occupational function, family care,
and self-image14.
The DASH utilizes a Likert scale, which presents a set of attitude statements
and then asks the respondent whether he or she agrees or disagrees and with
what degree of certainty according to a numerical scaling
system15. The DASH
has been validated by other investigators for the assessment of shoulder
disorders16. All
patients in the study completed the DASH questionnaire.
Patient expectations were evaluated with questions that are included in the
MODEMS. Preoperatively, the patients were asked: "What results do you
expect from your treatment?" with regard to six items: (1) relief from
symptoms, (2) to do more everyday household or yard activities, (3) to sleep
more comfortably, (4) to go back to my usual job, (5) to exercise and do
recreational activities, and (6) to prevent future disability (see Appendix).
The patients circled one response (ranging from "not at all
likely" [1 point] to "extremely likely" [5 points], or not
applicable) for each question, and an average score for the six questions was
then calculated for each patient. If a question was marked "not
applicable," the point value for that question was not included in the
average score. Postoperatively, the patients were asked: "Are the
results of your treatment what you expected?" regarding the first five
items listed above. Again, the patients circled one response
("definitely yes" [1 point], "probably yes" [2
points], "not sure" [3 points], "probably not" [4
points], or "definitely not" [5 points]), and the scores for the
five questions were averaged to derive a final score.
The MODEMS also includes the Short Form-36
(SF-36)17, a
thirty-six-item patient-administered questionnaire with eight scales: physical
function, role physical, bodily pain, general health, vitality, social
function, role emotional, and mental health. The SF-36, which evaluates the
general health status of a patient and can be completed in about ten minutes,
has been extensively tested for reliability and validity and has been used in
multiple studies to evaluate the effectiveness of treatment of orthopaedic
disorders18-20.
SF-36 raw scores were adjusted and reported as a percentage of expected values
for age and gender-matched controls.
The third section of the questionnaire included the Simple Shoulder Test
(SST), a shoulder-specific outcome measure that characterizes the functional
status of patients with an abnormal shoulder. The SST comprises twelve
yes-or-no questions about important activities of daily living that can be
performed by individuals with normally functioning shoulders. The questions
specifically assess shoulder comfort and
function21. The
ease of use, reproducibility, sensitivity, reliability, and responsiveness of
the SST have been documented for several shoulder
disorders22-24.
We report the results of the SST as the total number of positive
responses.
The fourth section of the questionnaire included visual analog scales for
rating shoulder pain, overall shoulder function, and overall quality of life.
The patients were asked to place a mark on a 10-cm line that was labeled
"none" on one end and "disabling" on the other for the
pain assessment, "comfortable" on one end and "can't use
it" on the other for the function assessment, and "little or no
problem" on one end and "very bad" on the other for the
quality-of-life assessment. After the patients marked each visual analog
scale, the mark was measured to the nearest millimeter and recorded as the
score.
Statistical Methods
Student t tests were performed to evaluate the significance of the
improvement in the outcome scores between the preoperative and postoperative
evaluations. Also, t tests were utilized to evaluate the difference between
patients who had been followed for one year and those who had not (and thus
had been excluded from the study) with regard to age, sex, whether they had
made a Workers' Compensation claim, the number of previous nonshoulder
surgical procedures, smoking, tear size, duration of symptoms, surgical
technique, and average patient expectation score. P values of <0.05 were
considered to be significant.
Multivariate regression analysis was used to evaluate the correlation
between outcomes and the number of comorbidities. In this analysis, patient
age, sex, whether the patient had a pending Workers' Compensation claim, the
number of any previous nonshoulder surgical procedures, smoking, tear size,
duration of symptoms, surgical technique (open, completely arthroscopic, or
mini-open), and average patient expectation score were used as confounding
variables. Stepwise regression analysis was performed by sequentially
eliminating insignificant confounding variables from the equation until the
final multivariate regression equation included all significant coefficients
along with the number of comorbidities. Comorbidity was included in the final
regression equation independent of significance. Regression coefficients and
significance levels were determined for each multivariate regression. P values
of <0.05 were considered to be significant. All multivariate regressions
included coefficients for each independent variable, which represent the
amount that each dependent variable changes with a one-unit increase in the
independent variable.
No significant differences (p > 0.05) were found between the patients
who had been followed for one year and those who had not with regard to age,
sex, number of medical comorbidities, smoking, number of previous nonshoulder
surgical procedures, patient expectations, tear size, or technique of
operative repair. Only Workers' Compensation status differed significantly
between the two groups: 31% of the patients who had been followed for one year
had a Workers' Compensation claim compared with 16% of those who had not been
followed for one year (p = 0.015).
In general, rotator cuff repair resulted in substantial improvements in
outcome. There were significant improvements in six of the eight subsections
of the SF-36 (physical function [p = 0.02], role physical [p < 0.001],
bodily pain [p < 0.001], vitality [p = 0.04], social function [p = 0.02],
and role emotional [p = 0.03]). Only mental health and general health
did not improve significantly (Table
I). There were also significant improvements (p < 0.001) in all
of the parameters of functional outcome, including the scores on the visual
analog scales for pain, function, and quality of life; the SST score; and the
DASH score (Table II).
There was a significant correlation between a higher number of
comorbidities and a significantly worse final postoperative score for four of
the eight SF-36 subsections (role emotional [p = 0.045], bodily pain [p =
0.032], general health [p = 0.001], and vitality [p = 0.033]) (see Appendix).
However, there was no significant correlation between the number of
comorbidities and the change between the preoperative and postoperative scores
for any of the SF-36 subsections (p > 0.05) (see Appendix). There was also
no significant correlation between the number of comorbidities and the final
postoperative pain, shoulder function, or quality of life as determined with
the SST, DASH, and visual analog scales (p > 0.05) (see Appendix).
Significant linear associations were demonstrated between the number of
comorbidities and the postoperative improvement in the visual analog score for
pain (p = 0.009), the visual analog score (p = 0.022) and DASH score (p =
0.044) for function, and the visual analog score for quality of life (p =
0.041) (see Appendix). Thus, patients with more comorbidities had a greater
improvement in the scores for shoulder pain, function, and quality of life
than did those with fewer comorbidities.
Several preoperative factors, including the size of the rotator cuff tear,
whether the patient had a Workers' Compensation claim, smoking, age, gender,
and duration of symptoms, have been evaluated to determine their effect on the
outcomes of rotator cuff
repair1-5,25.
In the orthopaedic literature, the effect of comorbidities on postoperative
outcomes has, to our knowledge, been evaluated only after total joint
replacement and after the surgical treatment of humeral nonunions. Dunbar et
al. found a significant negative correlation between comorbidities and the
scores on both general health measures (SF-36) and disease-specific outcome
instruments (Western Ontario and McMaster Universities Osteoarthritis Index
[WOMAC] and Oxford 12-Item Knee Score) after total knee
arthroplasty11.
Lingard et al. demonstrated similar results at two years after total knee
arthroplasties, with worse scores in the pain and functional domains of the
WOMAC scale and in the physical function domain of the SF-36 correlating with
a greater number of
comorbidities10.
Otsuka et al. demonstrated a negative effect of preoperative comorbidities on
postoperative SF-36 scores at a mean of forty-two months after surgical
treatment for humeral
nonunion26.
Although we found significant improvements on most of the subscales of the
SF-36, our results demonstrate significant negative linear associations
between preoperative comorbidities and one-year postoperative SF-36 scores for
role emotional, bodily pain, general health, and vitality. Consequently,
despite improvements in outcome related to shoulder pain and function,
comorbidities continue to have a profound effect on the general health status
of patients who undergo successful rotator cuff repair.
Our analysis of shoulder and upper-extremity-specific outcomes did not
demonstrate significant associations between the number of preoperative
comorbidities and the absolute postoperative functional outcome scores. In
contrast, we found a significant positive correlation between the number of
comorbidities and improvement in functional outcome scores. This observation
is counterintuitive in view of the findings in our previous study, in which
more comorbidities correlated with worse preoperative functional
scores7. The
findings of the current study suggest that rotator cuff repair restores
shoulder function to a point where the overall medical status of the patient
(number of comorbidities) no longer influences the self-reported shoulder
function. Perhaps this can be explained by the fact that patients with more
comorbidities start out with worse functional scores and therefore have more
room for improvement after surgical repair.
We believe that there are several possible explanations for the apparent
positive influence of comorbidities on the change in shoulder function. The
most likely is that the overall medical status of a patient (number of medical
comorbidities) is an amplifier of dysfunction and pain as perceived by
patients with a rotator cuff tear. Nevertheless, medical comorbidities
continue to have a profound negative effect on general health status, as
measured with the SF-36, after rotator cuff repair. In effect, medical
comorbidities do not have a positive influence on the outcome of rotator cuff
repair. Instead, at one year, comorbidities cease to have a negative
effect.
We recognize that this study has several potential and important
limitations. The lack of complete follow-up of all patients who had an
operatively treated rotator cuff tear raises the question of whether the
findings are valid for the entire population of patients with a rotator cuff
repair. Also, we did not assess rotator cuff healing (i.e., confirm it with
radiographic or arthroscopic examination) in our evaluation. Consequently, we
do not know the effect of a retear or failure of healing on our outcomes.
Healing of the rotator cuff has been shown to have a positive effect on the
final outcome after both open and arthroscopic
repairs27,28.
Finally, we assumed that each individual comorbidity had equal weight. While
we acknowledge that various comorbidities (e.g., cancer compared with high
blood pressure) are likely to differ in their effect on outcome variables,
weighting of the comorbidities would be arbitrary and inconsistent.
In summary, medical comorbidities do not appear to significantly affect
final shoulder function as assessed by the patient (with the DASH or a visual
analog scale) but do influence the final general health status after rotator
cuff repair. Therefore, a high number of medical comorbidities that do not
affect a patient's ability to safely undergo surgery should not be considered
a negative factor in determining whether a patient should have a rotator cuff
repair.
A table showing the questions used to measure patient expectations and
tables providing the results of the multivariate regression analyses are
available with the electronic versions of this article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?