Key words
gynecological operation - quality of life - recovery
Introduction
Convalescence is viewed as a process or as a time of recovery after illness or surgery
[1]. Postoperative recovery cannot simply be measured with individual parameters such
as freedom from pain. Instead, it is a complex multidimensional process which includes
social and economic elements in addition to physical and psychological aspects [2], [3]. To date, there are very few prospective studies which have investigated the influence
of several concomitant factors and their impact over time.
The time until patients can return to work after gynecological surgery for benign
indications is often relatively long [4]. Obviously, the invasiveness and complexity of the gynecological surgery, the patientʼs
expectations with regard to the duration of the recovery and the patientʼs preoperative,
functional state of health all have a predictive value on the length of time until
the patient can return to work; however, it appears that this time may also depend
on the recommended sick leave period [5], [6]. Recommendations by doctors about the appropriate length of time off work for patients
who undergo hysterectomy or another type of gynecological surgery vary considerably.
There are currently no evidence-based guidelines on the appropriate length of time
to be off work [7], [8]. The international data on benign diseases
is inconsistent. According to a data analysis carried out in Great Britain, the
recommended time off work after abdominal hysterectomy for benign indications ranges
from 2 to more than 12 weeks [9]. Johansen et al. reported in 2008 that less than half of women who undergo hysterectomy
returned to work again 1 week after the end of the recommended time off work [10]. Instead, the majority of patients extended their sick leave beyond the recommended
time on their own initiative [10].
Postoperative complications do not just negatively affect health-related quality of
life, they often also lead to an extended sick leave and are additionally associated
with poorer physical condition and persistent or intense feelings of anxiety [11], [12], [13].
To date, very few investigations have looked at whether sociodemographic factors also
have an impact on postoperative recovery after surgery. In their study, Brölmann et
al. found no significant association between patientsʼ level of education and the
time until returning to work after gynecological surgery [4]. However, unemployment was found to be a risk factor for developing depression in
patients who had recently undergone hysterectomy [14]. Similarly, low social support by the patientʼs partner, family or friends was associated
with a poorer outcome [15]. There are currently no larger systematic studies on postoperative recovery times
after gynecological surgery in Germany.
The aim of this prospective single-center study was therefore to find out whether
and how medical, psychological and/or sociodemographic factors affect postoperative
recovery.
Method and Patient Population
Method and Patient Population
Study design
All patients who underwent surgery for benign gynecological disease in the Gynecological
Department of the Charité Campus Virchow University Clinic in Berlin (Director: Prof.
Dr. Dr. h. c. J. Sehouli) over a 7-month period from June to December 2015 were consecutively
invited to participate in the study. Patients were interviewed at 4 different times
postoperatively: the first interview took place when patients were still in hospital
(T0: face-to-face interview); 3 subsequent interviews were carried out as telephone
interviews at 1 week (T1), 6 weeks (T2) and 7 – 8 months (T3) after being discharged
from hospital. All interviews (T0 bis T3) were carried out by the same study author
(S. S.). When carrying out interviews with Turkish-speaking patients, S. S. was assisted
by a trained student assistant with a Turkish migration background. Patients were
interviewed at the earliest on the first postoperative day and at the latest on the
day they were discharged from hospital.
Before patients were interviewed, they were provided with material informing
them about the research project and data protection. The questionnaire and the leaflets
were available in German and Turkish.
Inclusion criteria
Patients aged between 18 and 67 years who underwent surgery during the above-mentioned
period, gave their written consent to the study, and agreed to allow their phone number
to be recorded for the subsequent interviews T1 to T3 were included in the study.
Questionnaires and planned course of the study
The first questionnaire which was used for the in-hospital interview consisted of
4 parts (A – D). Part A collected sociodemographic data such as housing situation,
job or occupation, and school-leaving qualifications (for analysis, patients were
grouped into patients with a low level of education = no educational qualifications
or only attended primary school or a few years of secondary school; patients with
a moderate level of education = General Certificate of Secondary Education, or completed
secondary modern or middle school; patients with high level of education = school-leaving
examinations qualifying for admission to university or vocational baccalaureate or
university diploma/polytechnic degree). Information on the patientʼs satisfaction
with her current living situation, income, professional and family situation, health,
and general satisfaction with her life were also recorded; a Likert scale between
0 (“entirely dissatisfied”) and 10 (“entirely satisfied”)
was used to assess satisfaction. Part B of the questionnaire covered information
on migration and acculturation (Frankfurt Acculturation Scale [FRAKK]) [16]. Our study used the 2006 definition of “migration background” of Schenk et al.:
a person has a migration background if the person does not report German is their
first language or if the person was not born in Germany or has at least one parent
to whom this applies [17]. Part C consisted of an evaluated questionnaire on health (RAND-36). Part D asked
the patient about how she assessed her current capacity to work and about any feelings
of anxiety [18], [19].
In addition to the information obtained from the questionnaires, patientsʼ medical
data from their patient files were also used for data analysis.
In the 3 post-hospital telephone surveys T1 to T3, patients were always asked about
postoperative complications such as wound infections or secondary bleeding, whether
they currently required painkillers, whether they were currently able to work and
how strong their feeling of anxiety was (0 = “no anxiety” to 10 = “highest level of
anxiety” on an 11-point Likert scale; for the evaluation, “anxiety” was grouped into
“very little anxiety” = 0 – 3 points, “moderate anxiety” = 4 – 6 points, “strong anxiety” = 7 – 10
points). Patients evaluated their current capacity to work on a scale ranging from
0 = “incapable of work” to 10 = “fully able to work”. If the patient was not employed
or otherwise in work, she was asked how well she was able to carry out her daily responsibilities
(for example, household tasks). In addition, patients were asked the 10 questions
of the Recovery Index questionnaire [19]. During the second telephone interview (T2),
patients were additionally asked how long they were on sick leave or when they
would be able to fully return to work. The first day of sick leave was defined as
the first day after being discharged from hospital.
At the last post-hospital survey (T3) at 7 – 8 months after discharge from hospital,
patients were additionally asked again about their state of health using the RAND-36
questionnaire.
Statistical analysis of the questionnaires
The completed questionnaires were pseudonymized using a numerical sequence and analyzed
using SPSS (IBM SPSS Statistics Version 24).
Evaluation of the Recovery Index (RI)
Data analysis was based on the approach of Kluivers et al. [19]. A high RI score stood for a high level of recovery [19]. The Recovery Index was originally developed as a tool to survey patients who had
undergone hysterectomy [19] and was then used by Vonk Noordegraaf et al. to evaluate other patients who underwent
abdominal gynecological surgery [20]. As our study also included patients who underwent breast surgery, the original
text (“Even without doing any activity, I still regularly experience abdominal pain”)
was amended into a more general statement “… feel pain in the operated area”.
Evaluation of the RAND-36
The RAND-36 questionnaire of the RAND Corporation was used to record patientsʼ health-related
quality-of-life. This questionnaire corresponds to the questions of the Short Form
36 Health Survey questionnaire (SF-36) [18]. It consists of 36 questions which cover eight different domains: energy/exhaustion,
physical functioning, pain, perception of general health, physical role functioning,
emotional role functioning, social functioning, and psychological wellbeing. It also
includes a question about changes in health status over time [18]. The questions in the RAND-36 used at timepoint T0 all refer to the time immediately
prior to the surgical procedure and were then partially amended for use in the subsequent
surveys, for example, question 21: “How strong was your pain in the 4 weeks before the operation?”. The questionnaire was amended again for the interview at timepoint T3, with the
second
question of the RAND-36 questionnaire amended as follows: “Compared to the period before the operation how would you describe your current state of health?”. The RAND-36 questionnaire
was also used to investigate whether preoperative physical functioning or preoperative
psychological wellbeing had an impact on postoperative recovery. For this evaluation,
the scores for individual questions focusing on the domains “physical functioning”
and “psychological wellbeing” from the first interview (T0) were added up.
Classification according to level of invasiveness
As a first step, two of the authors (S. S., M. D.) classified the surgical interventions
into 4 levels of invasiveness similar to the classification used by Vonk Noordegraaf
et al. in their 2014 study:
-
“not very invasive”
-
“slightly invasive”
-
“moderately invasive”
-
“very invasive” [20].
As the case numbers in Group 1 were very low, they were grouped together with the
cases in Group 2 (“slightly invasive”), resulting in just 3 levels of invasiveness
([Table 1]). Some patients underwent combined procedures, with the individual procedures categorized
into different levels of invasiveness. In such cases, patients were classed according
to the most invasive procedure.
Table 1 Classification of surgical interventions into 3 levels of invasiveness (modified
from Vonk Noordegraaf et al. [2014] [20]).
|
Class 1 – slightly invasive:
|
-
Laser vaporization of condylomata on the vulva/vagina/ectocervix
-
Curettage for miscarriage
-
Hysteroscopy (with myoma/polyp ablation, if indicated)
-
Curettage
-
Loop conization
-
Removal of vaginal vault granulations
-
Abscess incision and drainage with placement of tamponade for breast/vulvar abscess
-
Marsupialization for Bartholinʼs cysts
-
Skinning vulvectomy
-
Suturing of vaginal margins to treat erosion in the vicinity of TVT sling
-
Excision of necrotic tissue and breast scar correction
|
|
Class 2 – moderately invasive:
|
-
Laparoscopy (myoma enucleation, resection of ovarian/paratubal/hydatid cysts, salpingotomy,
adenectomy, ovarian wedge resection, [partial] salpingectomy, resection of the uterosacral
ligament, hysterosacropexy, incision of pseudo-peritoneal cysts, coagulation of perforation
site following uterus perforation during curettage)
-
Laparoscopically assisted vaginal hysterectomy (LAVH) with/without adnexal surgery
-
Laparoscopically assisted supracervical hysterectomy (LASH)
-
Total laparoscopic hysterectomy (TLH) with adnexal surgery
-
Vaginal hysterectomy without adnexal surgery
-
Anterior and posterior colporrhaphy
-
Removal of breast implant and capsule with wound revision
-
Implant replacement and breast scar revision
|
|
Class 3 – very invasive:
|
-
(Mini-) laparotomy (myoma enucleation, hysterectomy with/without adnexal surgery,
supracervical hysterectomy, adenectomy, ovariectomy, resection of ovarian cysts, salpingotomy,
omentectomy, resection of peritoneal retention cysts, excision of parts of the uterine
wall to treat adenomyosis of the uterus, removal of necrotic tissue and suturing of
the uterine wall to treat wall dehiscence, adhesiolysis, chromopertubation)
-
Subcutaneous mastectomy with reconstruction
-
Mastopexy
-
Breast reconstruction with latissimus dorsi flap
-
Breast implant placement
|
The level of significance was set at p = 0.05. Internal consistency to measure the
reliability of the questionnaires used in the study was determined using Cronbachʼs
alpha coefficient (α), with SPSS used to calculate Cronbachʼs alpha coefficient. A
Cronbachʼs alpha coefficient of more than 0.80 indicates a good internal consistency
and therefore a reliable testing method [21].
Statistical methods used for evaluation
The investigated factors which could potentially have an impact included both nominally
and ordinally scaled variables as well as metric scaled measurements. Independent
variables were considered target values for the mean Recovery Index score, making
them dependent metrically scaled variables. All evaluated variables were combined
into a general linear mixed model to evaluate their impact on the course of the dependent
variable (i.e., the Recovery Index score) over time. A stepwise backward elimination
of the independent variables was carried out, in which the respective variable with
the highest p-value (highest redundancy) was excluded from the analysis. The value
of goodness-of-fit of the model was measured using the Akaike information criterion
(AIC). Based on this principle, independent variables are excluded from analysis until
the point when the goodness-of-fit of model, measured using the AIC, no longer improves.
Two of the research questions were examined with the help of univariate non-parametric
tests, as the underlying data did not meet the requirements for parametric tests.
Wilcoxon signed-rank test was used to compare the RAND-36 scores obtained at timepoint
T0 with those from timepoint T3. Mann-Whitney U-test was used to evaluate capacity
to work when comparing 2 independent groups, and Kruskal-Wallis H test was used to
compare 3 groups. If the latter test showed a significant result, a Bonferroni test
was carried out as a post-hoc test for pairwise comparisons of characteristics of
the independent variables.
Ethics vote and data protection
This research project was advised and approved by the Ethics Committee of Charité
Berlin, Ethikausschuss I, Campus Charité Mitte (application number EA1/111/15). The
study complies with the guidelines of Charité on ensuring good scientific practice
and with the Berlin Law on Data Protection.
Results
Study population
During the observation period from June to December 2015, a total of 269 patients
were asked to participate in the study during their stay in hospital. 81 patients
(30%) declined to participate. 57 of these patients (again 70%) had a migration background;
inadequate command of German meant that 34 of these 57 women were unable or unwilling
to participate in the study. Other reasons cited for declining to participate included,
in particular, the 3 planned subsequent telephone surveys. Some of the women felt
that the questionnaire was too big or that some of the questions were too personal.
Others felt that participating in the study was too overwhelming for psychological
reasons (e.g., because of their depressed mood postoperatively). Six of the 188 patients
who participated had to be subsequently excluded from in the study because the resected
specimen was found to be malignant. Ultimately, questionnaires from 182 women were
included at timepoint T0. At this timepoint
(T0), the surveyed women were between 19 and 67 years old (mean 40.45; SD: 11.56).
A total of 49 patients (27%) had a migration background.
Response rate for every timepoint of the survey
Out of the original 182 patients, 151 patients (38%) could be reached at timepoint
T1, 128 patients (70%) could be contacted at timepoint T2, and 106 patients (58%)
were reached at timepoint T3 for the respective telephone survey. The Recovery Index
(RI) questionnaire showed a good rate of reliability for all survey timepoints (T1-α:
0.870; T2-α: 0.813; T3-α: 0.813). This was also the case for the RAND-36 ([Table 5]).
Invasiveness of surgical intervention
A high Recovery Index (RI) score indicates a good recovery. As the level of invasiveness
of the procedure increased, the RI was only found to clearly decrease at timepoint
T1. Overall, the RI score increased over time ([Fig. 1]) irrespective of the level of invasiveness of the surgical procedure.
Fig. 1 Course of recovery (Recovery Index [RI]) over time.
Peri- and postoperative complications
Five patients (3%) had complications from surgery during their stay in hospital. The
patients who experienced complications during their stay in hospital reported significantly
lower RI scores. Similarly, RI scores were also lower if complications occurred postoperatively
in the time after discharge from hospital up until the third survey at timepoint T3.
Preoperative psychological wellbeing
The applicable domain of the RAND-36 was used to evaluate preoperative psychological
wellbeing. The mean score of the study population for this domain was 65.98 (SD: 18.13)
at timepoint T0.
Preoperative physical functioning
Out of a maximum possible score of 100, the mean score for this domain at timepoint
T0 was 85.16 (SD: 18.90) for all study participants.
Life satisfaction
During their stay in hospital, patients were asked about their level of satisfaction
(0 to a maximum of 10 points on a Likert scale) with regard to different aspects of
their life. The mean satisfaction score for all surveyed aspects of their life was
7.37 (SD: 1.38).
Postoperative feelings of anxiety
The mean score at timepoint T1 was 2.59 (SD: 2.64). 49 patients (33%) stated that
they had no feelings of anxiety, and only 4 patients (3%) reported a score of 9 points.
At timepoint T2 the mean score was 2.57 (SD: 2.54). 37 patients (29%) reported not
feeling anxious at all, while 2 women (2%) responded with a very high score, reporting
9 and 10 points, respectively. The mean score at timepoint T3 was 2.88 (SD: 2.40).
25 patients (24%) had no feelings of anxiety and 1 patient (1%) reported the highest
level of anxiety. The overall RI score decreases when patients report increasing levels
of anxiety.
The general linear mixed model, which included all previously described independent
variables scored at the different time points, confirmed that RI scores increased
significantly over time postoperatively ([Fig. 1]).
[Table 2] shows the results of this model, i.e. prior to backward elimination. To improve
the accuracy of the model, the following influencing factors were successively eliminated
from the full model: household, committed relationship, formal education, occupation,
migration background, satisfaction with life situation, level of invasiveness of the
surgical procedure, pre-existing conditions, and patient age. [Table 3] shows the results of the final model, i.e. the variables which had a significant
impact on the course of the RI after backward elimination of independent variables.
Sociodemographic variables such as migration background or level of education had
no significant impact. Evaluation of medical factors potentially affecting recovery
showed that in-hospital and postoperative complications, relevant previous operations,
and preoperative physical functioning had a significant impact on the recovery
process (RI). Of the evaluated psychological factors, postoperative feelings
of anxiety and preoperative psychological wellbeing had a significant impact on the
course of the RI, but sociodemographic factors did not. Moreover, overall recovery
(RI) improved over time ([Fig. 1]).
Table 2 Results of the evaluation based on a general linear mixed model prior to backward elimination, using the Recovery Index (RI) as the target value.
|
Influencing factors
|
p-value
|
F-value
|
|
Level of invasiveness
|
0.034
|
3.476
|
|
In-hospital complications
|
0.021
|
5.452
|
|
Pre-existing medical condition(s)
|
0.134
|
2.277
|
|
Previous operations
|
0.029
|
4.892
|
|
Living alone
|
0.231
|
1.448
|
|
School qualifications
|
0.715
|
0.337
|
|
In work
|
0.670
|
0.182
|
|
In a committed relationship
|
0.661
|
0.193
|
|
Migration background
|
0.152
|
1.907
|
|
Feelings of anxiety
|
0.026
|
3.700
|
|
Postoperative complications
|
< 0.001
|
77.773
|
|
Patient age
|
0.491
|
0.478
|
|
Satisfaction with life situation
|
0.194
|
1.709
|
|
Preoperative physical functioning
|
0.096
|
2.815
|
|
Preoperative psychological wellbeing
|
0.189
|
1.747
|
|
Time (improvement of RI over time)
|
< 0.001
|
82.559
|
Table 3 Additional statistics based on the general linear mixed model after backward elimination, using the Recovery Index (RI) as the target value.
|
Influencing factors
|
Coefficient
|
95% confidence interval
|
p-value
|
|
Complications occurred in hospital
|
|
|
|
|
|
− 7.935
|
− 13.226 – − 2.644
|
0.004
|
|
|
Reference
|
Reference
|
Reference
|
|
Previous operation(s)
|
|
|
|
|
|
− 2.138
|
− 3.751 – − 0.526
|
0.010
|
|
Feelings of anxiety
|
|
|
|
|
|
3.209
|
1.258 – 5.159
|
0.001
|
|
|
2.531
|
0.615 – 4.448
|
0.010
|
|
|
Reference
|
Reference
|
Reference
|
|
Postoperative complications
|
|
|
|
|
|
− 5.928
|
− 7.214 – − 4.642
|
< 0.001
|
|
|
Reference
|
Reference
|
Reference
|
|
Timepoint
|
|
|
|
|
|
− 6.662
|
− 7.863 – − 5.462
|
< 0.001
|
|
|
− 1.224
|
− 2.223 – − 0.225
|
0.017
|
|
|
Reference
|
Reference
|
Reference
|
|
Preoperative physical functioning
|
0.0635
|
0.0195 – 0.108
|
0.005
|
|
Preoperative psychological wellbeing
|
0.0587
|
0.0141 – 0.1033
|
0.010
|
|
Constant term
|
33.875
|
29.005 – 38.745
|
< 0.001
|
Health-related quality of life
The data were evaluated to see whether the scores of individual domains changed over
time. All scores, with the exception of the domains “emotional role functioning” and
“perception of general health”, improved significantly from timepoint T0 to timepoint
T3 ([Tables 4] and [5]).
Table 4 Descriptive statistics for individual RAND-36 domains at timepoints T0 and T3 (N = number,
SD = standard deviation, α = Cronbachʼs alpha).
|
RAND-36 domain
|
N
|
Mean
|
Median
|
SD
|
α
|
|
Physical functioning
|
|
|
|
|
|
|
|
182
|
85.16
|
90.00
|
18.90
|
0.880
|
|
|
106
|
90.24
|
100.00
|
15.86
|
0.876
|
|
Physical role functioning
|
|
|
|
|
|
|
|
182
|
58.24
|
75.00
|
43.51
|
0.903
|
|
|
106
|
83.49
|
100.00
|
32.35
|
0.896
|
|
Emotional role functioning
|
|
|
|
|
|
|
|
182
|
69.05
|
100.00
|
39.96
|
0.839
|
|
|
106
|
72.96
|
100.00
|
39.07
|
0.858
|
|
Energy/exhaustion
|
|
|
|
|
|
|
|
182
|
49.86
|
50.00
|
19.38
|
0.805
|
|
|
106
|
54.62
|
55.00
|
17.74
|
0.825
|
|
Psychological wellbeing
|
|
|
|
|
|
|
|
182
|
65.98
|
66.00
|
18.13
|
0.814
|
|
|
106
|
70.72
|
74.00
|
17.76
|
0.852
|
|
Social functioning
|
|
|
|
|
|
|
|
182
|
82.14
|
100.00
|
26.26
|
0.874
|
|
|
106
|
92.10
|
100.00
|
17.28
|
0.794
|
|
Pain
|
|
|
|
|
|
|
|
182
|
65.38
|
67.50
|
33.50
|
0.878
|
|
|
106
|
90.47
|
100.00
|
20.67
|
0.906
|
|
Perception of general state of health
|
|
|
|
|
|
|
|
182
|
65.11
|
70.00
|
21.33
|
0.767
|
|
|
106
|
66.27
|
72.50
|
24.17
|
0.854
|
|
Changes to state of health
|
|
|
|
|
|
|
|
182
|
42.31
|
50.00
|
24.42
|
–
|
|
|
106
|
66.04
|
50.00
|
27.64
|
|
Table 5 Comparison between the individual domains of the RAND-36 survey for timepoints T0
and T3 – results of significance testing.
|
RAND-36 domain
|
p-value
|
|
Physical functioning
|
0.001
|
|
Physical role functioning
|
< 0.001
|
|
Emotional role functioning
|
0.129
|
|
Energy/exhaustion
|
0.026
|
|
Psychological wellbeing
|
0.008
|
|
Social functioning
|
0.001
|
|
Pain
|
< 0.001
|
|
Perception of general state of health
|
0.326
|
|
Changes to state of health
|
< 0.001
|
Duration of time signed off work
Six weeks after discharge from hospital, patients who were employed or in training
were asked about the length of time they were signed off work and patients who were
self-employed or freelancers were asked how long it had been until they were fully
able to work again. These questions applied to 132 patients. The remaining women in
the study were either not working or had already retired at the time of the survey.
Of these 132 women, 95 patients (i.e., 72%) were surveyed again and the length of
time they were off work was recorded. All medical sick notes were issued by registered
physicians. The mean time off work was 24.08 days (SD: 15.32). [Fig. 2] shows the time signed of work as a function of the level of invasiveness of the
procedure. As expected, the higher the level of invasiveness, the longer the time
off work.
Fig. 2 Time signed off work in days, according to the 3 levels of invasiveness (n).
Estimation of fitness for work
As described above, patients were asked at each subsequent telephone interview how
they evaluated their current fitness for work using a scale ranging from 0 = “incapable
of work” to 10 = “fully able to work”. At timepoint T1 the mean patient score was
5.22 (SD: 3.11), at timepoint T2 it was 7.87 (SD: 2.46), and at timepoint T3 it was
8.52 (SD: 2.10). Patientsʼ fitness for work increased continually over time.
Scores were subsequently compared to see whether postoperative fitness for work changed
according to the level of invasiveness of the procedure and whether the occurrence
of postoperative complications had an impact on patientsʼ estimation of their fitness
for work. At timepoint T1, the estimation of patientsʼ fitness for work differed significantly
between strongly and moderately invasive procedures and between strongly and slightly
invasive procedures (p < 0.001 and p = 0.002, respectively). At timepoint T2 only
the difference between very and moderately invasive was significant (p = 0.028). At
time T3 there was no significant difference in the estimation of fitness for work
relating to the invasiveness of the procedure.
Analysis showed, however, that the occurrence of postoperative complications led to
a significantly poorer estimation of fitness for work at all 3 timepoints (T1: p < 0.001,
T2: p < 0.001, T3: p = 0.018).
Discussion
Prospective studies on the factors influencing postoperative recovery which are based
on complex statistical analysis and investigate changes over time are rare. Most previous
studies which looked at recovery after gynecological procedures in the widest sense
of the term compared the effects of different hysterectomy techniques. In one study
published in 2008, Persson et al. compared recovery times for laparascopic and abdominal
hysterectomies, and in a second study from 2010 they compared recovery times after
supracervical and total abdominal hysterectomy and found no differences in postoperative
recovery [12], [22]. In contrast, according to the data by Brummer et al., laparoscopic hysterectomy
had the shortest stay in hospital and the shortest time off work compared to vaginal
or abdominal procedures [23]. The study by Vonk Noordegraaf et al. on which we partly based the idea
for our study compares different gynecological abdominal procedures and divided
them into 4 different classes according to their level of invasiveness [20]. Vonk Noordegraaf et al. found that increasingly levels of invasiveness of surgical
procedures were associated with an increased risk of longer times off work [20]. Another study has shown that the time until patients can resume everyday activities
becomes longer when the invasiveness of the procedure increases [24]. Similarly, the patients in our study who underwent the most invasive procedures
were off work longest.
Our results about the recovery process were unexpected. Contrary to prior assumptions,
more invasive procedures were not associated with delayed recovery. Other factors
such as prior operations or pre-existing medical conditions played a greater role
in the recovery process and led to a significantly poorer course of recovery. According
to a study by Dessources and colleagues, patients with specific pre-existing medical
conditions or whose pre-existing medical conditions meant that they had a higher ASA
classification were more likely to require readmission to hospital after the initial
discharge [25]. According to a study by Theunissen et al., ASA classification is also a predictor
of postoperative recovery: patients with ASA classification level III had a poorer
recovery than patients with ASA class I [26].
Postoperative complications are associated with longer times off work, a higher rate
of readmissions to hospital, and a higher risk of developing chronic postoperative
pain [13], [22], [25], [26]. Moreover, postoperative complications can result in a reduced capacity to work
and higher feelings of anxiety [11]. Our study found that strong feelings of anxiety postoperatively had a negative
impact on recovery. According to a study by Kagan et al., feelings of anxiety which
are already present preoperatively have a negative impact on postoperative recovery
[27]. According to other studies, this can lead to higher levels of postoperative pain
or even result in chronic postoperative pain [28], [29]. Our study did not
investigate what the impact of complications which occurred in prior surgical
procedures was on recovery. This point will be examined in further planned study.
An earlier study by Reitsma et al. showed an improvement in quality of life postoperatively
after gynecological surgery [30]. The studies published to date have primarily focused on women who underwent hysterectomy.
The studies mainly reported improvements in physical wellbeing, but often also found
a positive impact on patientsʼ psychological situation and social relations [31], [32], [33]. Only one study reported that hysterectomy had no relevant impact on the patientsʼ
psychological state [34].
The patients surveyed in our study who were in work were signed off work for a mean
of 24.1 days (the time off work after slightly invasive procedures was 15 days; the
time off work after moderately invasive procedures was 20 days, and the mean time
signed off work after very invasive procedures was 34 days). The time off work after
slightly invasive procedures appears to be relatively long. However, similar figures
have been reported elsewhere, for example in the study by Brummer et al. which compared
the effects of different hysterectomy procedures [23]. Women who had an abdominal hysterectomy were off work for 32 days, women who had
a vaginal hysterectomy were off work for 30 days, and women who underwent laparoscopic
hysterectomy were off work for 22 days after surgery [23]. In the 2014 study by Vonk Noordegraaf et al., the time until returning to work
was even longer, with 60 days off work following
moderately invasive and 69 days off work after very invasive surgical procedures;
the time off work after slightly invasive procedures was 14 days [20]. Brölmann et al. reported in their study that 8 weeks after a gynecological operation,
only 50% of patients had fully returned to work, 30% of patients were working part
of the time by this point, and 20% were still signed off work [4].
However, it should be noted that it is not possible, based on the number of days on
which women are signed off work with a medical sick note, to differentiate between
patient-related (not robust enough yet, etc.) and physician-related factors (unclear
basis for determining duration of sick leave, subjective influences culminating in
an extension of sick leave to accommodate patient wishes, and similar). Moller et
al. noted that gynecologists who issued sick notes for women undergoing hysterectomy
gave longer periods of sick leave to women whose work involved heavy physical work
compared to women who only had to carry out light physical work [8]. There are some general indications that patients who are self-employed and therefore
unable to earn any income in the time they are off work tend to return to work earlier
than employees, although not all studies have borne this out [5], [20], [35]. Sometimes sick leave is extended beyond the recommended period at the request of
the affected women [10]. In their study, Bouwsma et al. found that both the period until patients returned
to work and the period until various physical activities were resumed were longer
than the period recommended by experts [24]. Moreover, patient expectations of having a long period of sick leave can lead to
an extended sick leave [20]. These factors could have resulted in patients in our study being on sick leave
for a relatively long time after not very invasive procedures, even though the expected
physical limitations and pain after such operations are usually low(er). Because of
the resulting time away from work, this also has an economic relevance for society
as a whole.
One week after being discharged from hospital, women without postoperative complications
had recovered around 60% of their capacity to work and after six weeks they had recovered
80% of their capacity to work. After 7 – 8 months, patients had recovered a mean of
90% of their capacity to work. Women who experienced postoperative complications reported
a significantly lower capacity to work at all 3 timepoints (T1: 30%, T2: 60%, T3:
70%). These figures are comparable with the findings in the study of Doll et al. [11].
Strengths and Limitations of the Study
Strengths and Limitations of the Study
Strengths
To our knowledge, this is the first study on this topic in Germany which measures
changes over time using 4 timepoints spaced at relatively long intervals, uses internationally
validated questionnaires and expressly includes women with a migration background.
Limitations
-
It was not possible to question all of the patients over the entire period of the
study. At timepoint T3 only 58% of the women originally enrolled in the study were
still available for questioning.
-
Postoperative recovery is a complex, multidimensional process. It is possible that
our questionnaire did not cover all aspects.
-
The study was a single-center study.
Conclusions for Clinical Practice
Conclusions for Clinical Practice
The development and implementation of so-called Fast Track Surgery (FTS) or Enhanced
Recovery After Surgery (ERAS) programs to reduce time in hospital, complication rates,
the rates of readmission to hospital, and hospital costs are now quite common [36], [37], [38]. Such programs mainly focus on the immediate postoperative phase when patients are
still in hospital.
Vonk Noordegraaf et al. have developed an e-health program for the Netherlands specifically
for gynecological patients [39], [40]. In this program, women who underwent gynecological surgery for benign indications
were given personalized recommendations with regard to resuming everyday activities
and returning to work. The aim was to make it easier to reintegrate at work [39], [40]. Such programs along with the necessary scientific monitoring and evaluation should
be developed and implemented in Germany to particularly support those patients who
are at risk of having longer postoperative recovery times. With the help of further
studies based on our results, it should be possible to identify patients who are at
risk. In addition, such programs could help reduce the time off work, particularly
after less invasive procedures which our study found to be
associated with relatively long sick leave times, by offering targeted information
and support.
Note
This study was supported by a research grant from the German Society for Psychosomatic
Gynecology and Obstetrics (DGPFG).
