Health-related quality of life after pelvic organ prolapse surgery in Ethiopia: A 6 month follow-up study

Pelvic organ prolapse (POP) affects health-related quality of life (HRQoL). Patient-reported outcomes (PRO) measures the quality of care from the patient’s perspective. PROs are an important measure of surgical outcome and used to calculate health gains after surgical treatment. We assessed HRQoL in women undergoing surgical repair of POP. Two hundred fteen women with stage III or IV prolapse underwent surgical POP repair between February 2018 and May 2019. Pelvic Organ Prolapse Quality of Life (P-QoL-20) was administered at baseline, 3 and 6 months postoperatively to assess HRQoL. Depressive symptoms and body image also evaluated. Linear mixed-effect models were used to compare pre and postoperative HRQoL scores and investigate potential predictors.


Results
Participant's mean age was 49.3 ± 9.4 years. Most (81.9%) had stage III prolapse and underwent a vaginal hysterectomy, although 40% preferred uterine preservation. No differences were seen between women follow-up and those lost to follow-up in HRQoL (p > 0.05). The P-QoL, depressive symptoms, and body image were improved 6 months post-operatively. The change in P-QoL signi cantly associated with body image scores. The type of surgery did not show a signi cant difference. Being married showed an improvement in the personal relationship subscale score (β = 5.8, p < 0.01).

Conclusions
Our results indicated a potential improvement of HRQoL after surgical treatment. The result could be useful for patient counseling on the expected HRQoL outcomes of surgical treatment. Surgical service should be accessible for women suffering from POP to improve HRQoL.

Background
Pelvic organ prolapse (POP) is a common problem affecting health-related quality of life (HRQoL) [1,2]. It is one of the most prevalent gynecological conditions requiring surgical treatment with a lifetime risk of 11-19% based on data from High Income Countries (HIC) [3,4]. The surgical repair of POP seems to improve HRQoL [5], however, it is not always possible because of anatomic failure, recurrence of symptoms or postoperative complications [6,7]. Vaginal hysterectomy, for instance, has been found to be inadequate for restoring pelvic support [8]. A recent review also reported subjective and objective recurrence rates of 7-14% and 23-41%, respectively [9]. Furthermore, there are signi cant cost implications for POP surgery after failure of the rst surgery [5,10].
The de nition of surgical success following POP repair is debatable between patient and surgeon [11,12]. In studies of POP surgery, objective and satisfaction outcomes, and HRQoL are recommended to be reported [13]. However, most studies have focused on anatomical success without considering subjective outcomes [5]. The contemporary de nition includes not only the anatomical success but also patientreported outcome measures (PROMs) [13].
In most low and middle-income countries (LMICs) surgery is expensive, facing catastrophic expenditure paying for surgical care [14]. Because of the high cost to patients, it is imperative that surgical care is of high quality and has a meaningful impact on patients. But little evidence of PRO use in LMICs including Ethiopia [15]. In Ethiopia, unlike obstetric stula, there is no free surgical service to POP. Yet, the Ethiopian

Ministry of Health in collaboration with the United Nations Population Fund and Women and Health
Alliance International regularly organizes the " POP surgical campaign" in governmental hospitals to treat symptomatic women [16]. A recent study in Ethiopia reported 46% symptomatic POP requiring treatment [17].
The HRQoL is a multi-dimensional concept that evaluates the perceptions of the impact of disease or treatment on physical, psychological, and social function [18]. It has been used as a primary outcome measure to evaluate the effectiveness of POP treatments [19]. The primary goal of POP surgery is to provide quality care with meaningful patient impact, i.e. reduce symptoms and improve HRQoL [12].
However, in HIC, outcomes of studies on HRQoL are inconsistent [20,21] and mostly focused on anatomical outcomes [5]. In LMICs including Ethiopia, PRO outcomes after POP surgery were infrequently reported. Recent qualitative ndings from Ethiopia reported that rural women receiving free-of cost POP surgery experienced great bene t in many aspects of their life [22]. Although the outcomes of surgery on HRQoL is hardly comparable between women living in high versus low income contexts, more evidence on the effect of POP surgery are needed speci cally in low-income contexts. Therefore, this study aimed to assess HRQoL before and after POP surgery via the use of PRO tool, the Prolapse Quality of Life.

Methods
A single-group longitudinal study (pre-post) was conducted on women admitted at the University of Gondar referral hospital (UoGH) for prolapse repair from February 2018 to May 2019. The study time period ended in November 2019 with the 6 months follow-up of the last study subjects. Women were eligible for inclusion if they had symptomatic stage III or IV prolapse according to the Pelvic Organ Prolapse Quantitation (POP-Q) classi cation system [23] and desired surgical correction. Women who had surgery performed by an abdominal approach, with current co-morbidities, or who had previously undergone POP surgery were excluded.

Surgical Procedures
The surgical treatment consisted of the correction of all the defects. The decisions as to which technique to use and which defects to treat were left to the discretion of the operating surgeon. Cystocele and rectocele were treated with conventional anterior and posterior colporrhaphy. For uterine descent, either vaginal hysterectomy (VH) or uterine-preserving procedures (UPP, such as sacrospinous xation) were performed. All procedures were performed under spinal anesthesia. During discharge, patients were informed to avoid sexual intercourse and heavy lifting/work load for approximately 3 months and asked to have a follow-up visit at 3 and 6 months.

Sample Size Calculation
Assuming a paired difference of 3, the standard deviation of the differences 15 [24], considering α = 0.05, power of 80% and a possible loss of 15% during the follow-up, sample size became 224.

Study Instruments
The following instruments were used to investigate PROs: Prolapse Quality of Life (P-QoL), POP Symptom Score (POP-SS), Body Image in POP (BIPOP) and Patient Health Questionnaire (PHQ-9) to evaluate HRQoL, presence and/or severity of POP symptoms, body image perception, and depressive symptoms, respectively.  [25]. The Amharic version had three components: physical (PC; including GHP, PI, PL, RL SL, and SM), psychological (PSC; including EMO and SE), and social/personal relationship component (PRC; containing PR). The Amharic version is easy to use, can be administered face-to-face and is publicly available [26]. This study measured QoL using domains for PC, PSC, and PRC of the P-QoL instrument.
The POP-SS comprises 7-items and scored from 0-28, with higher scores are indicative of more bothersome symptoms [27]. The POP-SS used in the present study was validated in the Amharic language, and details found elsewhere [28].
The PHQ contains 9-items, with higher scores indicating more severe depressive symptoms [29]. It has been validated for use in primary care and obstetrics/gynecology outpatient clinics to diagnose major depressive disorders [30]. The current study utilized the Amharic version [31].
The BIPOP includes 10-items regarding the possible impact of POP on a woman's perception of her body including genital body image, sense of attractiveness and risk for sexual dysfunction [32]. Each item uses a 5-point Likert response with lower scores indicating better body image [33]. The English version was translated into Amharic and back to English to con rm the correctness.

Data Collection Methods
Data were collected preoperatively, 3 and 6 months postoperatively by trained female nurses. All baseline interviews were administered face-to-face at UoGH and subsequent interviews were conducted either at participant's home, nearby clinics or UoGH. Preoperatively socio-demographic data including age (years), residence (urban or rural) marital status (partner or not), employment status, and educational level (literate or not) were collected.
Data such as the stage of POP, duration of POP symptoms (years) and type of operation (VH or UPP) were also abstracted from patient medical records using a standardized form. Furthermore, women were asked to complete all the PROs including P-QoL, PHQ-9, BIPOP, and POP-SS. These same PRO instruments were administered postoperatively. Duration of POP symptoms denotes the number of years from the time when prolapse symptoms (feeling/seeing of vaginal bulging) rst occurred and then classi ed as delayed to seek healthcare if she stayed more than a year [34].

Statistical Analyses
Each completed instrument was checked visually for completeness before fed to a computer. Data were summarized using mean with standard deviation (SD) or median with interquartile ranges (IQRs) for continuous, and numbers with percentages for categorical variables.
The primary outcome was change in P-QoL scores (9 domains, PC, PSC, and PRC). Linear mixed effect models were used to assess the change in QoL. Furthermore, linear mixed-effect model was used to investigate the potential predictors. First, P-QoL outcome measurements (PC, PSC and PRC subscales) over time were compared using a random intercept model assuming time as a xed effect. Then, the models were tted with PC, PSC and PRC scores as dependent variables, and time, age, type of surgery, POP duration, marital status, POP-SS, PHQ, and BIPOP scores as a covariate ( xed effect). For each model, we reported the xed effects coe cients (β value) of the independent variable with the associated 95% CI and P-value. A model with random intercepts and slopes and having an unstructured covariance structure was employed after model comparison with the Akaike information criterion (AIC). The unstructured covariance structure, which accounts for the within-subject correlation, was chosen based on the model t using AIC. For all analyses, p < 0.05 was considered statistically signi cant and normality was assessed using Shapiro-Wilk test. Statistical analyses were performed using STATA, version 14.0. A lower mean P-QoL scores, indicating improvement, were expected postoperatively.

Ethical approval
The study protocol was approved by the University of Gondar Institutional Health Research Ethics Review Committee (UoG (O/V/P/RCS/05/216/2017). Written informed consent was obtained from all individual participants included in the study.

Response rate
Of the 226 subjects enrolled, 215 (97.7%) took part in the baseline interview. Eleven patients revoked their consent before the operation. At 3 months postoperatively, data was recorded for 193 (89.7%) subjects.
At the 6 months follow-up, 185 (86.0%) were assessed, leaving 30 patients dropped out. The primary reasons for leaving the follow-up schedule included declined further participation (n = 8), lost to follow-up for unknown reasons (n = 8), relocation (n = 10), died for reasons unconnected to the study (n = 2), or incomplete data (n = 2).

Socio-demographic characteristics
Mean patient age at the time of surgery was 49.3 ± 9.4 years (35 − 70 years). Mean number of children born to the patients was 5.9 ± 2.6 children. One-fth had under-ve children. The majority patients had stage III prolapse (81.9%), and more than half underwent a vaginal hysterectomy. However, almost twofth of them would prefer uterine preservation preoperatively, provided that outcomes were equal (Table 1). Reasons reported as the main barriers to seek early treatment were lack of money (21.2%), fear of disclosure (15.9%), the perception that POP is incurable (13.6%), fear of treatment outcome (12.1%), lack of accompanying support (8.6%), distance from a health facility (requiring 2 days or more to reach health facility) (6.2%), and lack of transportation (5.0%). However, 59.5% had decision making power to visit health care when getting sick (data not shown).

Quality of life, body image and depressive symptoms
A signi cant improvement in the score of PRO instruments (P-QoL, POP-SS, BIPOP, and PHQ) were recorded over the course of the study (p < 0.001, Fig. 1, Table 2). There was variability in the score at baseline, and 3 and 6 months postoperatively (Fig. 2). , and the impact on emotions (64.4 ± 27.5) was also experienced preoperatively. A signi cant improvement was recorded after 6 months for the above listed domains as well as for other investigated domains. All domains of P-QoL improved signi cantly after 6 months of surgery compared to their status prior to operation ( Fig. 1 and Table 2). This positive effect on QoL remained stable in patients surveyed at 3 and 6 months. However this improvement did not reproduced in personal relationships domain score during this follow-up period, and patient had worse score (15.0 ± 12.6 to 20.0 ± 18.3, p = 0.005, paired t-test). Quality of life was signi cantly higher as measured by PC, PSC and PRC scores at 6 months (p < .001, Fig. 2 and Table 2).
Patients' prolapse related symptom scores dropped gradually from the baseline to the 6 months in the longitudinal analysis. POP-SS score signi cantly dropped from 16.6 ± 4.3 to 2.1 ± 1.2. Likewise, patients' depressive symptom scores dropped from the baseline (16.4 ± 6.5) to the 6 months (7.8 ± 4.3) after surgery. Furthermore, we observed an improvement in BIPOP score over time, indicating a better body image perception after surgery. The BIPOP score reduced from 33.9 ± 7.3 to 17.4 ± 5.6 (p < 0.001, Fig. 2 and Table 2).

Predictors of quality of life
LMM analysis was performed to examine the longitudinal effects of sociodemographic, POP-SS, BIPOP and PHQ on P-QoL domains/components ( Table 3). The change in P-QoL after surgery was associated with the change in POP-SS, PHQ and BIPOP scores (p < 0.001). There was a change on PC and PRC score overtime for a point change in the POP-SS, PHQ and BIPOP scores. Each improvement in the POP-SS score resulted in an approximately 7 and 13 point improvement in the PC and PRC scores, respectively (p < 0.001). Being married resulted in a 5.7 point increase in the PRC score (p < 0.01) ( Table 3).  Age, parity, type of surgery and prolapse stage were not associated with improvement of QoL scores.

Discussion
In this study surgical repair of POP effectively improve POP symptoms, body image and HRQoL. This adds to the growing body of literature that POP and its surgical treatment is associated with HRQoL.
In order to ensure the quality of surgical outcome, reliable, valid and easy-to-use measures of surgical quality and patient impact are needed. The P-QoL is a PRO tool that can measure the impact of surgical interventions on women's HRQoL. This study uses P-QoL to demonstrate an improvement in patientreported HRQoL after surgery in Ethiopia.
PRO tool assesses patient's opinion thereby measure a patient's health status or HRQoL. Usually PRO data are collected through short and self-completed questionnaires. But many are di cult to administer and may be di cult for populations with high illiteracy rates to understand [1]. However, P-QoL is simple and quick to use tool approximately taking 30 min to administer using face-to-face [26]. Our study adds to the evidence that P-QoL can be used as a PRO tool to demonstrate patient impact after surgery in LMIC, speci cally Ethiopia.
The 86% follow-up rate observed in our study is much higher compared with Nepal's study that reported 64% [35].
Follow-up rates in LMICs are typically lower than those in high-income countries (HICs) although even follow-up rates vary considerably in HICs [36,37]. In this study, the decline in further participation, lost to follow-up for unknown reasons and relocation were the reasons for leaving the follow-up schedule. To minimize the loss of patients to follow-up, a trained data collector periodically checked the data integrity and dealt with missing data by acquiring information from patients and family via repeated phone or inperson interviews. In addition, we also compared the baseline pro les of those who completed the study and those who dropped out. However, there was no difference in bassline socio-demographic, depressive and POP symptoms, body image and POP stage associated with P-QoL scores between the two groups (p > 0.05, Table 4).
In the current study surgical repair of POP showed a signi cant improvement of QoL for all 9 investigated domains of P-QoL instrument over the course of the study. Patients had a signi cantly lower impairment of their QoL at 3 months after surgery. At 6 months after the procedure, the QoL of the study participants had improved even more, and the difference to the preoperative gures remained signi cant. The better QoL scores compared to baseline scores were also observed in all the 3 P-QoL components (PC, PSC and PRC) after 6 months. Previous studies in Western countries using P-QoL instrument also con rmed that correction of POP via alloplastic vaginal meshes can signi cantly improve patients' QoL [24,[38][39][40][41][42].
However, patients bene ted from a highly signi cant improvement in all domains of QoL, although the surgery was native tissue repair in current study. Surgeries using the patient's own tissues are the rst-line operations in the treatment of POP [43]. Our nding also compare favorably with previous studies in LMICs, although a different QoL tool used [35]. Furthermore, qualitative ndings from Ethiopia reported that rural women receiving free-of cost POP surgery experienced great bene t in many aspects of their life [22].
In our study, the average scores for prolapse effects on physical and role activities and personal relationships was quiet high at baseline. A previous study carried out in European women also showed similar scores for the same domains [24,38,39]. Daily life for Ethiopian rural women (e.g. the burden of physical activities/work, gender inequality affecting personal relationships, shame, lack of education/knowledge etc.) is hardly comparable to women living in Europe, and thus the higher baseline [22].
The remarkable improvement found in P-QoL domain scores (PC and PRC) after surgery could have a positive in uence on formerly impaired HRQoL among those affected. This is similar to a study from Nepal [35], reporting a signi cant improvement in every aspect of the QoL measured. In rural settings where nearly all the housework was preformed by women alone or by help of their children and even women often help out on heavy farming activities [44,45], the improvement of PC after surgery enabled them to perform daily household and/or outdoor roles like fetching water from distant, participating in farmining activities and help in caring for under-ve children.
Patients referred for surgery frequently complain about bulging and associated urinary, bowel or sexual symptoms, which are responsible for a signi cant decrease in HRQoL [46]. In the current study prolapse symptoms, especially vaginal bulging, and lower abdominal heaviness signi cantly improved postoperatively. Furthermore, the overall total score of POP-SS was signi cantly reduced after surgery.
Similar results have been reported elsewhere [35,36,42]. This observation of improvement may be reassuring to the patient and clinician and may very well motivate more women suffering from POP to seek help.
The negative impact of POP on body image and sexual life has been well documented [47][48][49]. The current study demonstrated a signi cant improvement in body image 6 months after surgery. Similarly, another study found better body image and sexual satisfaction after surgical intervention [36,50]. The patient may consider the altered genital anatomy due to the surgery to have a signi cant impact on their general sense of attractiveness. Moreover, the improvement of body image may be due to reduction of POP symptoms. Evidence showed strong association of POP symptoms and body image score [47][48][49].
In our study, worse body image score was reported preoperatively.
Despite 40% of women preferred uterine preserving procedures provided that outcomes were equal, more than half of them underwent a VH in this study. Although difference was not observed among groups in HRQoL scores, women's preference has to be considered for better surgical outcomes [51]. A more recent study in Ethiopia reported an abdominal approach with mesh augmentation as the most successful surgical repair for POP [52].
Prolapse symptoms are not life threatening but has signi cant negative effects on mental health of woman [15,53]. Our study demonstrates that depressive symptoms improved signi cantly after surgical treatment. This ndings is consistent with the results of previous studies showing signi cant improvement in the score of psychological and social health and QoL after surgical treatment [42,54,55].
Prolapse may increase symptoms of depression, while symptoms of depression may impact health behavior, symptom burden, QoL, and functional impairment pre-and post-operatively [56]. Our research found that surgery leads to a dramatic improvement not only in condition-speci c QoL but also in depressive symptoms. Improved mental health status (PHQ and PSC score) could have positive effect on women's QoL after surgical intervention [15,53].
Marital status had a signi cant association with the change in PRC domain score. Those who were married had greater improvement in HRQoL score than counterparts. This nding is supported by a qualitative study in Ethiopia, which reported that women who lived alone experienced poor improvement in their lives after surgery. For these women life continued to be a struggle [22]. They also found that avoidance of returning to heavy chores shortly after surgery depended substantially on the support from their family and community members, and proved di cult for those living alone [22]. This might be because the probability of getting social or relative support is better for those who live in marital bonds.
In the current study there was no difference in improvement in POP speci c QoL following surgery by age group, parity, and stage of POP. Similar result on age group was also reported elsewhere [57].

Strengths and limitations
This study does have limitations. One is that with the 6 months follow-up period, our ndings may only allow assessment of short to medium-term HRQoL outcomes. A long-term follow-up is needed to draw rm conclusions with regard to HRQoL and anatomical outcomes. Another limitation is that although our sample was women with POP, the study design was limited to single setting and lack of control group.
Including women who had POP without intervention as a control group is unethical. Moreover, to evaluate the effect of surgery on HRQoL it would require a control group of similar patients without surgical repair.
Since HRQoL among women post-surgery may not be comparable to women with an intact uterus, irrespective of their POP status, we did not include a comparison group to evaluate changes in HRQoL.
Hence, our ndings could theoretically be biased in directions that are di cult to predict and we cannot completely rule out the possible effect of changes occurring over time. The single setting warrants that the applied surgical technique was the same for every patient, and, therefore, the nal outcome could be compared. A third limitation is that although we have no reason to doubt the truthfulness of the responses given form respondents, it is conceivable that patients may have withheld less socially desirable responses. The free surgical services received may also affect a patient's willingness to report a negative outcome. Furthermore, the patient survey based on questions about their QoL, body image, depressive symptoms are limited as the statements were obtained from the PRO instrument survey. The nal limitation is losses to follow-up. Given that subjects who follow-up were not statistically different from lost to follow-up, results did not substantially affect and may be generalizable to the entire surgical population in Dabat district. A signi cant bias may occur even with a small proportions of patients lost to follow-up [58], despite more than 20% poses serious threats to validity in general [59].
Despite the above limitations, our study has a number of strengths. This is one of the rst studies to report on HRQoL among women underwent POP surgery in Ethiopia. Although POP is a complex and sensitive topic to study, the use of a disease-speci c validated PRO tool, i.e. P-QoL, is an important strength. This enabled us to evaluate women's POP symptoms, body image, and HRQoL in a local context. The use of these instruments further made the study ndings internationally comparable. The follow-up rate of 84% is relatively high and the numbers are large compared to other follow up studies in LMICs.

Conclusions
In this study, we focus on patient-centered results in evaluating surgical outcome. Our study shows that there was a signi cant improvement in patient-reported HRQoL, body image, and depressive symptoms after surgical repair of POP in Ethiopian women. We recommend that P-QoL is used as a PRO tool by those involved in the delivery of surgical care to monitor the patient-centered impact of surgical interventions. Access to surgical services for disadvantaged women may also be important to improve HRQoL in women with POP.

Declarations
Con ict of interest.
All authors declare that they have no con ict of interest.

Availability of data and materials
The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.
Funding. Quality of life before and after surgical repair of prolapse. The bar chart shows the gures for QoL at 3 and 6 months after surgery compared to patients' QoL before surgery, where 100 corresponds to the lowest QoL. The QoL was itemized into a various domains. Changes in scores between the baseline and the 3 and 6 months follow-up period were all signi cant (p < 0.001).