Evaluation of the SF‑36 questionnaire for assessment of the quality of life of endometriosis patients undergoing treatment: A systematic review and meta‑analysis

Endometriosis is a condition resulting from ectopic endometrial cell implants that grow progressively and determine an inflammatory response ( 1 ). The ovaries are the most common locations for endometriotic lesion formation, followed by the anterior and posterior cul-de-sac, the posterior surface of the broad ligaments, the uterosacral ligaments, the serosa of the uterus, the fallopian tubes, the sigmoid colon, the appendix, and the round ligament ( 2 ). The therapeutic options are based on clinical presentation, including infertility, pain, or mass; they are individualized according to disease severity, patient age, and reproductive desires while considering the medication side effects, costs and morbidity of the surgical procedures ( 3 , 4 ). Endometriosis negatively affects the physical, psychological, and social aspects of women's lives ( 5 ), and therefore has a great an impact on the health-related quality of life (HRQoL). It is difficult to characterize the quality of life (QOL) among specific patient groups because it is a broad concept that includes life satisfaction, good health, education, personal and family safety, adequate housing, employment, interrelationships, and leisure pursuits ( 6 ). Globally, in the scientific literature, QOL is evaluated using a multitude of validated and non-validated questionnaires, such as the 36-item survey generic questionnaire (SF-36). It includes eight conceptual health domains: General Health (GH), Physical Functioning (PF), Bodily Pain (BP), Role Physical (RP), Vitality (VT), Social Functioning (SF), Mental Health (MH), and Role Emotional (RE); these are summarized in the physical (PCS) and mental (MCS) component scores ( 7 ). Recently, a number of original studies and systematic reviews have evaluated the impact of endometriosis on patient QOL ( 8-10 ). Studies have also reported the impact of different therapeutic approaches on the QOL of patients with endometriosis ( 11 ). A systematic review detailed the quality of sexual life after endometriosis-targeted laparoscopic surgery ( 12 ). Another recent meta-analysis showed that endometriosis-targeted surgical resection can improve the major domains of QOL ( 13 ). However, systematic reviews that evaluate the impact of surgical and non-surgical treatments on the QOL of patients with endometriosis are scarce. The main objective of the present review was to investigate the efficiency of questionnaire SF-36 in assessing the QOL of patients with endometriosis who were undergoing medical or surgical treatment.

We utilized the ‘Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)’ guideline to conduct this study’ ( 14 ). A combination of key word terms including ‘endometriosis’, ‘quality of life’ and ‘questionnaire SF-36’ were utilized for the literature research conducted in October 2018. The utilized databases included PubMed/Medline, Embase, and Cochrane; some other additional sources and references of the identified articles were also utilized. The search strategy that was utilized for PubMed/MEDLINE is outlined as follows. Articles that evaluated the quality of life (QOL) and used the SF-36 to assess women with endometriosis were evaluated. Articles that interviewed patients both before and after endometriosis-related surgical or medical treatment were included in our analysis. We also utilized the reports of women with endometriosis that were compared with a control group. Studies that focused on any type of endometriosis and were published in English, German, or French were considered in our analysis. There were no restrictions based on publishing year, journal, author, or institution. All the selected studies were independently evaluated by two reviewers. Any disagreements regarding the articles were solved by the third more experienced author and the senior authors. The data were extracted in tables that consisted of first author's name, publication year, study type, number of initial participants, treatment type, presence of a control group or lack thereof, endometriosis type, treatment, follow-up, and the final number of participants. To evaluate the quality of the included studies, the modified Newcastle-Ottawa Scale (NOS) was applied ( 15 ). This scale helped to assess the risk of bias in the remaining studies. The studies were classified as ‘good’ when they fulfilled at least 70% of the 12 criteria, ‘fair’ when they fulfilled at least 50%, and ‘poor’ when they fulfilled less than 50%. Since we observed between-study heterogeneity, we applied random-effects meta-analyses, using odds ratios (OR) and 95% confidence intervals (95% CI). The results were presented as pooled prevalence values. Studies that reported data by comparing various endometriosis types were included in the data for the most severe, whereas studies that dealt with different types of surgical treatment were considered the most difficult interventions. Statistical analysis was performed using the Open-Meta analyst [CEBM Brown, open-source, OpenMetaAnalyst for Windows 10 (64-bit)]. In our study, statistical results were calculated as follows. The first report was based on a comparison between patients with endometriosis and a control group (healthy women); the second considered surgical treatment, and the third group considered hormonal treatment.

Study research. In accordance with our research plan, we initially identified the records of 657 articles that corresponded to our search strategy. Furthermore, 25 articles were sourced generated from additional sources (cross references). Any duplicates were removed, and the final database consisted of 345 articles. Excluding some articles due to language issues, we obtained reviews, case reports, or abstracts from 127 records for screening. Among these, 55 studies were excluded based on the inclusion criteria. Finally, full-text evaluation was conducted for 72 studies. The qualitative analysis was based on 37 articles, whereas the quantitative analysis utilized 14 articles. Regarding the study types, our qualitative analyses included 21 prospective studies, 3 retrospective studies, and 7 randomized clinical trials. A total of 25 studies that associated other questionnaires with the SF-36 to evaluate the QOL of patients with endometriosis were identified. Based on our study protocol, we observed that 13 studies did not apply the SF-36 questionnaire before and after treatment. The results were specified only in the graphics and not in the text. Some studies reported the results as a total number; thus, it was impossible to include them in the statistics. Only studies that had all the reported numerical data from the questionnaires were selected for the meta-analyses. Based on the inclusion criteria, we obtained 37 studies; among these, 16 articles were considered as being of ‘good’ quality (between 6 and 8 points), and 11 articles were considered as being of ‘fair’ quality (between 3 and 5 points). Table I includes evaluations for each study. The characteristics for each study are recorded in Table II . The research included 1,1101 women, among whom 6,888 patients were diagnosed with endometriosis. Based on the endometriosis type, 17 studies dealt with all types of endometriosis, 9 studies focused on deep infiltrative endometriosis, and another 9 studies identified bowel endometriosis or deep infiltrative endometriosis with bowel involvement. QOL was evaluated by using only the SF-36 in 12 studies including 1,912 women and by using SF-36 associated with various other questionnaires in 25 studies that had a collective total of 8,022 women. As part of the SF-36 questionnaire's general evaluation, it was applied for studies that presented the QOL of the patients with endometriosis in comparison to that of healthy women or patients with endometriosis that were interviewed using SF-36 just once (without any follow-up period). Based on this classification, the review included 15 studies with a general overview on endometriosis. Among them, 6 studies evaluated the QOL of patients with endometriosis compared with that of healthy women, and 9 studies compared different types of treatment among patients with endometriosis. On this topic, 6,810 women were evaluated, and 6,393 had endometriosis. The classification based on endometriosis subtypes confirmed that 5,410 patients had all the endometriosis types, 416 had deep infiltrative endometriosis (DIE), and 198 had minimal endometriosis. For this group of the review, the quantitative analysis could be performed in 6 studies. Our study revealed that patients with endometriosis had an altered QOL compared to the control healthy women group. The pooled prevalence showed that the selected parameters were influenced as follows: Physical Functioning (PF): OR, 78.87 (95% CI, 68.97-88.77; I 2 =98.77%; P≤0.001); Bodily Pain (BP): OR, 51.63 (95% CI, 45.23-58.03; I 2 =92.62%; P≤0.001); Role Physical (RP): OR, 56.12 (95% CI, 45.08-67.12; I 2 =95.40%; P≤0.001); Vitality (VT): OR, 43.09 (95% CI, 32.23-53.95; I 2 =98.67%; P≤0.001); Social Functioning (SF): OR, 58.10 (95% CI, 47.28-68.91; I 2 =97.67%; P≤0.001); Mental Health (MH): OR, 55.65 (95% CI, 45.58-65.72; I 2 =98.41%; P≤0.001); and Role Emotional (RE): OR, 57.84 (95% CI, 47.21-68.48; I 2 =95.19%; P≤0.001). Based on poled prevalence, for patients with endometriosis, the most affected parameter was Physical Functioning (PF): OR, 78.87 (95% CI, 68.97-88.77; I 2 =98.77%; P≤0.001). Qualitative analysis regarding QOL for patients that underwent surgical treatment for endometriosis was evaluated with a SF-36 tool, and it included 21 studies with a collective population of 3,368 patients ( Table III ). In all the selected studies, the questionnaire was completed before and after surgery. The various studies' follow-up periods ranged between 7 weeks and 3 years. The types of endometriosis were all types of endometriosis (3 studies), DIE (14 studies), or intestinal endometriosis (3 studies). Statistical analyses were performed for 5 studies that included mean values for each result. The studies confirmed that the parameters evaluated with SF-36 improved after surgical treatment. Pooled prevalence revealed that each parameter improved as follows: Physical Functioning (PF): OR, 63.39 (95% CI, 48.71-78.07; I 2 =97.65%; P≤0.001); General Health (GH): OR, -0.71 (95% CI, -1.20--0.23; I 2 =90.04%; P≤0.001); Bodily Pain (BP): OR, 41.25 (95% CI, 38.95-43.55; I 2 =97.33%; P≤0.001), Role Physical (RP): OR, 43.16 (95% CI, 32.74-53.58; I 2 =91.83%; P≤0.001); Vitality (VT): OR, 37.99 (95% CI, 30.54-45.44; I 2 =93.45%; P≤0.001); Social Functioning (SF): OR, 48.09 (95% CI, 30.67-63.50; I 2 =98.69%; P≤0.001); Mental Health (MH): OR, 51.78 (95% CI, 44.70-58.87; I 2 =92.73%; P≤0.001); and Role Emotional (RE): OR, 50.65 (95% CI, 41.37-59.93; I 2 =88.39%; P≤0.001). In the case of patients who had their endometriosis surgically treated, among the parameters evaluated by SF-36, the highest rate of improvement was observed in Physical Functioning (PF): OR, 63.39 (95% CI, 48.71-78.07; I 2 =97.65%; P≤0.001). Quantitative analysis was conducted for 7 studies on patients who underwent hormone treatments, which were evaluated with SF-36, whereas the statistic forest plots were based on 3 studies. The patients with endometriosis (392 in number) were compared with a control group that did not receive the same treatment (placebo or control). The studies showed that parameters evaluated with SF-36 improved after hormonal treatment. Pooled prevalence revealed that each parameter improved as follows: Physical Functioning (PF): OR, 38.65 (95% CI, 14.39-62.91; I 2 =38.65%; P≤0.001); General Health (GH): OR, -0.60 (95% CI, -1.78-0.57; I 2 =94.13%; P≤0.001); Bodily Pain (BP): OR, 28.51 (95% CI, 10.05-41.96; I 2 =97.33; P≤0.001); Role Physical (RP): OR, 32.42 (95% CI, 0.62-64.49; I 2 =98.62%; P≤0.001); Vitality (VT): OR, 31.49 (95% CI, 6.58-56.39; I 2 =98.58%; P≤0.001); Social Functioning (SF): OR, 35.86 (95% CI, 12.01-59.7; I 2 =97.92%; P≤0.001); Mental Health (MH): OR, 35.11 (95% CI, 5.72-64.50; I 2 =98.85%; P≤0.001); and Role Emotional (RE): OR, -0.39 (95% CI, -0.90-0.20; I 2 =74.57%; P≤0.001). After the hormonal treatment, the parameter that showed the most improvement was Physical Functioning (PF): OR, 38.65 (95% CI, 14.39-62.91; I 2 =38.65%; P≤0.001).

Our meta-analyses revealed that the SF-36 questionnaire had a heterogenous applicability in all the studies. Recently, Chauvet et al performed a recent systematic review, which proved that the SF-36 was the most frequently used scale, followed by the EHP-30, for evaluating quality of life (QOL) of patients with endometriosis ( 16 ). The design of our study was different from previous meta-analyses as we included articles that evaluated patients with endometriosis using the SF-36 questionnaire before and after surgical or medical treatment or as a comparison tool between groups. The literature did not reveal a similar approach. For example, Arcoverde et al performed another meta-analysis utilizing generic questionnaires and revealed that surgery improved the quality of life for patients with all types of endometriosis ( 17 ). We excluded cases with hormonal approach and cases that were evaluated using different questionnaires ( 18 ). There was no meta-analyses that evaluated the efficiency of SF-36 for patients with endometriosis treated with hormonal therapy or as a comparison method between study groups. Regarding the first issue of our classification, we identified 6,393 women who received treatment for endometriosis and answered the SF-36 questionnaire. Based on this approach, we observed that, for women with endometriosis, compared with the control groups, physical functioning was the main parameter that received the most alteration. Two peculiar studies were included within this group of studies. Laursen et al ( 19 ) compared patients with fibromyalgia/whiplash, endometriosis, lower back pain, or rheumatoid arthritis and healthy women. We included this study because of its data accuracy; furthermore, there was no other pathological influence on the women with endometriosis. The second study was the Verket et al study ( 20 ), which evaluated women from the general population, women with endometriosis, and women with rheumatoid arthritis. We incorporated it into the analyses of this study because it did not have a control group of healthy women for comparison to women with endometriosis, and the arthritis had no impact. For both of these studies, we considered only the data for the endometriosis patients. Van Aken et al demonstrated that, for endometriosis patients, pain cognition and QOL are independently associated ( 18 ). Verket et al highlighted that patients with moderate to severe endometriosis had overall impaired QOL compared to women from the general population ( 20 ). Women with DIE reported constipation, defecation pain, appetite disorders, longer evacuation time, and increased stool consistency without laxatives ( 21 ). Neuromuscular electro-stimulation was found to be an efficient method for alleviating endometriosis-related pain ( 22 ). The present analysis of QOL for patients with endometriosis who received surgical treatment included data from 3,368 patients. The pre- and post-surgery questionnaire in all the studies showed that physical functioning was the principal parameter that improved after the interventions. The studies proved that endometriosis symptoms such as pain showed improvement after surgical treatment ( 23 ), even though endometriosis also influenced biopsychosocial variables ( 24 ). Patients with minimal endometriosis seldom reported surgery as a good treatment for improving QOL ( 25 ). For patients with bowel endometriosis, QOL improved significantly after laparoscopic treatment ( 26 , 27 ), with long-term results ( 28-30 ). The study by Angioni et al ( 31 ), a randomized clinical trial, included a surgery group as well as a hormonal treatment group. For this group, we considered the arm A results, with a complete resection of DIE. The study by Roman et al (2018) ( 32 ) was analyzed with regard to the segmental resection of the rectum, and the study by Ledu et al ( 33 ) was utilized for the group consisting of patients with stoma; this was because of the rectal resection for DIE with rectal involvement (the study results were relevant for this group). Some authors have recommended surgery for rectosigmoid endometriosis in cases where the disease is unresponsive to conservative treatments ( 34-36 ). Daraï et al supported the finding that osteopathic manipulative therapy has a good impact on endometriosis symptoms in women with colorectal involvement ( 37 ). However, although the radical surgical approach can have a major positive impact on QOL, it does not improve all symptoms, as reported by Garavaglia et al ( 38 ). De Graaff et al concluded that many patients with endometriosis experienced better QOL and continued to have endometriosis-associated symptoms even when they were treated in tertiary care centers ( 39 ). Roman et al reported that conservative surgery should be reserved for patients with large deep rectal endometriosis ( 32 ). Daraï et al recommended laparoscopy for patients that required radical en bloc hysterectomy and colorectal resection ( 40 ). Laparoscopic colorectal resection was also indicated for cases of lower intraoperative and postoperative complications for patients with rectal endometriosis due to higher spontaneous post-surgery pregnancy rates ( 41 ). Araujo et al reported that gonadotropin-releasing hormone agonist (GnRHa) administration was followed by a temporary improvement in pain among patients with incomplete resection of deep infiltrative endometriosis (DIE), but it did not affect postsurgical pain when DIE implants were completely excised ( 42 ). The study by Angioni et al ( 31 ) was considered for the medical treatment group because it applied postoperative randomization (of initial groups with complete or incomplete resection of endometriosis) to receivers or non-receivers of GnRHa. It selected a group with complete resection of endometriosis, which received GnRH, and compared it with complete resection of endometriosis without medical treatment. Regarding statistics, the Caruso group ( 28 ) was not considered because it compared GnRH receivers with AINS receivers. Mabrouk et al confirmed the role of combined oral contraceptive therapy in curbing the progression of dysmenorrhea and dyspareunia and the growth of deep endometriotic nodules ( 43 ). Meanwhile, Miller concluded that the stimulatory phase of the GnRHa therapy could be correlated with an increase in pain and a decrease in the QOL ( 44 ). Márki et al emphasized that pain management and emotion regulation strategies can be utilized for improving the QOL of patients with endometriosis. The authors suggest that it could be obtained through simultaneous application with physical treatment and psychological care ( 45 ). Zhao et al conducted the first study to explore only the effects of psychosomatic therapy on the QOL of endometriosis patients and demonstrated that progressive muscular relaxing (PMR) training could serve as a reference for future psychosomatic interventions regarding endometriosis. PRM works efficiently for anxiety and depression related to endometriosis ( 46 ). Melis et al further suggested that a patient-centered approach that is integrated with various medical, psychological, and sexual issues should be applied in the treatment of patients with endometriosis ( 47 ). Nunes et al compared the QOL of patients with fibromyalgia and women with and without endometriosis. The QOL of women with endometriosis showed the highest impairment ( 48 ). Teixeira et al demonstrated that potentated estrogens significantly reduced endometriosis-associated pain compared to a placebo ( 49 ). Sesti et al demonstrated that, for endometriosis stage III-IV, QOL can be improved through postoperative hormonal suppression treatment or dietary therapy compared to surgery as well as a placebo for relieving pain ( 50 ). Regardless of the differences across country- and ethnicity-level contexts, endometriosis was found to impair QOL and work productivity for women who continued to experience diagnostic delays in primary care ( 51 ). Even if hypocortisolism could be sometimes considered a consequence of the aversive symptoms of endometriosis, the etiopathology remains unelucidated ( 52 , 53 ). The relationship between endometriosis and ovarian cancer was studied by Grandi et al and the preventive effect of inhibiting ovulation and menstruation was highlighted ( 54 ). EHP-30 and SF-36 questionnaires are sensitive to changes related to endometriosis, especially correlations between increased pressure pain sensitivity and hyperalgesia and the impairment of QOL parameters, as reported by Laursen et al ( 19 ). Thus, healthcare resource allocators should focus on women with endometriosis as the disease may have a different impact for each woman ( 55 ). The SF-36 seems to be a valid instrument for evaluating QOL in women with endometriosis ( 56 , 57 ). In conclusion, the 36-item survey generic questionnaire, SF-36, seems to represent an efficient/valid method for assessing the QOL of patients with endometriosis who are undergoing surgical or hormonal treatments. It can be applied before and after procedures or in order to compare study groups.