Health-related quality of life among cervical cancer survivors at a tertiary hospital in Kumasi, Ghana

Background: Cervical cancer survivors often experience signicant diminution in health-related quality of life (HRQoL). We aimed to investigate the overall HRQoL, determine the role of the stage of disease and type of treatment received on HRQoL, and evaluate predictors of HROoL among cervical cancer survivors in Kumasi, Ghana. Methods: A hospital-based cross-sectional study was conducted in 153 cervical cancer patients who completed curative treatment between January 2004 and December 2018 at Komfo Anokye Teaching Hospital. The European Organization for Research and Treatment of Cancer core questionnaire (EORTC QLQ C-30) supplemented with the cervical cancer-specic (EORTC QLQ-CX24) module was used. The Kruskal-Wallis test was used to determine the effect of the stage of cervical cancer and the type of treatment received on mean scores of the different domains of HRQoL. Multivariate logistic regression was performed to identify predictors of HRQoL. P < 0.05 was considered statistically signicant. Results: The mean global health score (GHS) was 79.7 (+/-16.2), and it differed signicantly with International Federation of Gynaecology and Obstetrics (FIGO) stage, with 84.1 76.2, and 79.9, for stages I, II and III respectively (p=0.012). Financial burden was higher in participants with FIGO stage II compared to stage I disease (45.4 vs 20.8, p=0.012.). The mean GHS scores for surgery, chemoradiation and radiation only were 85.2, 75.9, and 82.1 respectively (p=0.027). Compared to participants who were treated with Chemoradiation, those treated with surgery had lower nancial diculties (12.1 vs 41.6 , p=0.019), better body image (95.7vs 79.5, p =0.039) and better symptom experience (5.9 vs 12.0, p =0.043).The likelihood that survivors’ GHS is affected is decreased with illiteracy (AOR = 0.30, 95% CI=0.09–1.00), An assessment of the impact of patient- and treatment- related factors on the survivor’s quality of life (QoL) will provide insight into whether the survival benets are justiable, when compared with limitations posed by patients and treatment characteristics. We aimed to investigate the overall HRQoL, HRQoL by stage and treatment received, and predictors of HRQoL among cervical cancer survivors in a single centre in Kumasi, Ghana. Results can help set targets for planners, treatment benchmarks for healthcare providers and create national benchmarks for care that correlate with improved quality of life for survivors. modality impact on their GHS, followed by survivors who received radiotherapy alone. The GHS, body image, nancial diculties and symptoms reported by survivors differed with the treatment option. In this study, surgery was associated with the highest GHS/overall QoL, which is consistent with earlier studies. 17,20 Earlier studies also reported more symptoms in women treated with radiotherapy with or without chemotherapy. 33 The symptoms experienced after treatment may underline the signicant alteration in body image observed across the various treatment modalities. The loss of reproductive organs and external scarring of the genitalia via radical surgery and radiotherapy respectively, have been reported to negatively impact survivors’ psychophysical identity and cause distortion of the body image. 33 The absence of uterus and cervix, and the inability to reproduce traditionally, although often overlooked, is a major source of stress or discomfort to young women living with or surviving cancer. 34 These organs are seen by survivors as embodiment of womanhood, femininity and fertility. In most of the cultures in Sub-Saharan Africa, childbearing is a signicant expression of femininity.

Educating women about expected impacts on their quality of life, and strategies to manage and mitigate these impacts, is essential.

Background
Although the incidence of cervical cancer (CC) remains low in many countries, it is signi cantly higher in developing countries. 1,2 Cervical cancer is the second most common cancer among women in Ghana. Treatment of CC is primarily by surgery or radiation therapy, with chemotherapy as an important adjunct. 3 In Ghana, only 5.2% of women with CC present with Stage I disease, and are likely to bene t from surgery. 4 The risk of nodal involvement in Stage IA1 disease is low, and surgical treatment by conization and simple hysterectomy is often adequate. 3 Approximately, 7.4 % of patients with stage IA2 disease have pelvic lymph node involvement (PLNI), and modi ed radical hysterectomy and pelvic lymphadenectomy (PLND), with or without bilateral salpingo-ophorectomy, is standard treatment. Cone biopsy and radical trachelectomy plus PLND are options for women who desire future fertility. 5,6 Radiotherapy may also be used to provide comparable treatment outcomes for early invasive disease in women with severe comorbidities, such as morbid obesity with attendant detrimental anaesthetic risk, uncontrolled hypertension, or diabetes mellitus. 3 In the majority of low and middle income countries (LMICs) including Ghana, most patients present at a later stage where radiotherapy is the treatment of choice. 4,7 The treatment of locally advanced disease (FIGO Stage IB3-IVA) has changed over the years from radiotherapy alone to concurrent chemoradiotherapy, with a number of studies demonstrating survival advantage of cisplatin-based chemotherapy in both adjuvant and de nitive setting. [8][9][10] Cervical cancer survivors often experience physical symptoms and psychosocial distress that adversely affect their health-related quality of life (HRQoL). HRQoL encompasses domains of life which are directly affected by the presence of disease or treatments. 11 Several generic and disease-speci c tools have been developed for the assessment of HRQoL. These are multidimensional, patient-reported tools, which assess aspects of life that are directly affected by changes in health. 11,12 he European Organisation for Research and Treatment of Cancer (EORTC) HRQoL tool is a commonly used questionnaire to assess quality of life among all cancer patients (e.g. EORTC QLQ C30), and can be supplemented by the EORTC-QLQ CX-24 to focus speci cally on cervical cancer patients or survivors. 12,13 In high-resource settings, the focus of cancer care has shifted from traditional treatment outcome measures (e.g. 5-year survival outcomes) to patients' symptoms and quality of life as they progress through treatment and into post-treatment life with cancer as a chronic disease. 13 However, in lowresource settings like Ghana, there is limited data on HRQOL among cervical cancer patients and survivors. An assessment of the impact of patient-and treatment-related factors on the survivor's quality of life (QoL) will provide insight into whether the survival bene ts are justi able, when compared with limitations posed by patients and treatment characteristics. We aimed to investigate the overall HRQoL, HRQoL by stage and treatment received, and predictors of HRQoL among cervical cancer survivors in a single centre in Kumasi, Ghana. Results can help set targets for planners, treatment benchmarks for healthcare providers and create national benchmarks for care that correlate with improved quality of life for survivors.

Study setting
The study was conducted at Komfo Anokye Teaching Hospital (KATH), the second largest teaching hospital in Ghana. It is located in Kumasi, the capital city of Shanti Region. The hospital was established in 1942 and has more than 1000 beds providing diagnostic and treatment services for about 260,000 to 300,000 patients per year. The logistics and human resource base for gynaecologic oncologic services have developed through international cooperation and perform full multi-disciplinary care for gynaecologic cancer patients. 14 The Gynaecologic Oncology Unit and Department of Oncology provide services for over 400 women with genital cancer annually. Ethical approval was obtained from the Committee on Human Research, Publications and Ethics (CHRPE) of the School of Medical Sciences, Kwame Nkrumah University of Science and Technology and KATH (CHRPE/AP/661/19) Written and verbal informed consent were obtained for participants interviewed via face-to-face interaction and telephone respectively.

Study design and participants
A hospital-based cross-sectional study was conducted from July to September 2019. Participants were de ned as women diagnosed with cervical cancer who completed curative treatment between 2004 and December 2018, at Unit of Gynaecologic Oncology and the Department of Radiation Oncology of KATH. A total of 462 records of women with disease-free status after curative treatment were identi ed ( Figure 1). An estimated sample size of 145 had adequate power to detect a global health score with a standard deviation of 21.5 and a desired margin of error of 3.5 scores. 15 Considering a 10% of contingency for inappropriate and nonresponses, computer-generated simple random sampling without replacement was used to select 160 participants. A total of 153 women were willing to participate in the study and none of the patient data were incomplete. Participants were recruited consecutively until the required sample size was reached. Survivors were considered eligible if they were disease-free at least 6 months prior to the study. Women who were treated for malignancy of another anatomic site or who were critically ill were excluded from the study.
Data was collected by two trained oncology nurses working within the oncology clinic. Two days training was given for the oncology nurses focusing on; the contents of the questionnaire, the identi cation of patients based on the inclusion/exclusion criteria, and how to get consent. Participants were assured of anonymity and con dentiality of their information obtained in the study by excluding any personal identi er in the data collection form. They were also reassured that the report of the ndings would not identify them and only the aggregate data would be reported. Telephone interviews or face-to-face interviews were conducted in Twi, a local language, or in English to administer the questionnaires. Telephone interview were conducted for patients who lived outside Kumasi city or were not scheduled/did not report for review within the study. Socio-demographic and clinical characteristics of the participants were extracted from their medical records. All the collected data were checked for completeness by the principal investigator on a daily basis.

Study variables
Dependent variables: global health score (GHS), functional scales and symptom scales Independent variables: Socio-demographic characteristics such as age, time since completion of treatment, level of education, occupation, marital status, residence, proximity to treatment facility and tribe.
Clinical and treatment characteristics such as body mass index (BMI), parity, FIGO stage, primary treatment type, level of haemoglobin, units of blood transfused, and other comorbid conditions.

Instrument and scoring
The evaluation of the HRQoL was performed using the EORTC core questionnaire (EORTC QLQ C-30), supplemented with the cervical cancer-speci c module EORTC QLQ-CX24. 16 The 30-item EORTC QLQ-C30 is the most frequently used HRQoL assessment tool. Data on the use of the tool in Ghana is scarce, but it has been validated in many geographic and cultural settings. 15,17 . It is comprised of 15 subscales, 5 functional domains (physical, role, cognitive, emotional and social functioning); 3 multi-item symptom scales (fatigue, pain, and nausea/vomiting), 6 single items that assess additional symptoms commonly reported by cancer patients (dyspneoa, appetite loss, sleep disturbance, constipation, and diarrheoa), perceived nancial di culties and GHS. The EORTC QLQ-CX24 is a 24-item scale, grouped into 3 multiitem scales, 11 items with symptom experience domain, 3 items with body image domain, and 4 items with sexual/vaginal functioning domain. Further, it has single-item scales, that assess lymphoedema, peripheral neuropathy, menopausal symptom, sexual worry, sexual activity, and sexual enjoyment. All scores on the EORTC QLQ-C30 and QLQ-CX24 were transformed into 0 to 100 scale according to the EORTC QLQ scoring manual. 18 A higher score on the GHS and functioning scale represented a better level of functioning. A higher score on the symptom scale represented worse levels of symptoms/ high level of problems or ill-health.

Data Analysis and interpretation
Descriptive statistics were used to present the QoL scores, and patient-and treatment -related variables. The Kruskal-Wallis test was used to determine the effect of stage of disease and type of treatment received on mean score of the different domains of HRQoL. The HRQoL domains were then dichotomized; functional domain and global health score/QoL score < 75 were considered affected (75 or above score indicated "no problem at all" ) whereas on the symptom scale, a score 25 and above were considered affected (below 25 indicated " no problem at all"). 19 Stepwise multivariate logistic regression was performed to identify predictors of HRQoL. Only variables with p < 0.25 and clinically signi cant variables were included in the multivariable logistic regression model. All hypotheses involved were twosided tests; p < 0.05 was considered statistically signi cant. An internal consistency of the multi-item scales was calculated using the Cronbach a coe cient; values of a >= 0.70 were considered acceptable for group comparisons. All analyses were performed using Stata 13 (StataCorp, College Station, TX, USA).

Sociodemographic, clinical and treatment characteristics of cervical cancer survivors
A total of 153 women completed the questionnaires and seven declined participation on personal reasons. Nine survivors were critically ill ( Figure 1). The mean (+SD) age of the survivors was 58.3 (+ 11.4) years, with a median follow up time of 33.8 months (range: 6.3 to 180.5 months) after completion of treatment. (Table 1). Time since completion of treatment was more than one year in the majority (86.3%) with 29.4% being long-term survivors. Regarding level of education, 61 (39.9%) had no formal education and (11.1%) received senior high school or tertiary education. About fty-seven percent had anaemia, necessitating haemotransfusion in a third (32.0%) of the cases, before or during treatment ( Table 2). Most participants had FIGO stage III cervical cancer (37.3%). Primary chemoradiation was the treatment option in about half (47.7%) of the women. Twenty-eight (18.3%) participants underwent primary radical surgery with or without adjuvant radiotherapy.

Health-related quality of life (HRQoL) of cervical cancer survivors
The mean GHS (±SD) score was 79.7 (±16.2). The EORTC QLQ-C30 multiple and single-item scales were also calculated and the mean (±SD) score of the functional scales ranged from 88.2 (±14.1) to 91.8 (±18.4), with the highest being social functioning and the lowest being cognitive functioning (Table 3). For the symptom scales, the mean (±SD) nancial di culty score was the highest 32.8 (±40.9) followed by pain 17.9 (±24.0) and fatigue 14.0 (±19.0). The EORTC QLQ-CX24 domains mean (±SD) score ranged from 50.0 (±40.7) for sexual enjoyment to 85.3 (±26.0) for body image. On the cervical cancer-speci c symptom scales, the lowest mean score 7.0 (±19.7) was for lymphoedema and the highest was sexual Mean score differences of EORTC QLQ-C30 and QLQ-CX24 scales with stage and treatment characteristics On the EORTC QLQ-C30 scale, only GHS/Overall QoL and nancial di culty score differed signi cantly with regards to the FIGO stage and primary treatment type. (Tables 4). Participants who survived with stage I disease had a better GHS/Overall QoL score with lower nancial di culties compared to those who had FIGO stage II and III disease. More women with Stage II disease received chemoradiation. On the cervical cancer speci c module, body image and symptoms differed signi cantly with respect to the treatment modality applied, with surgical treatment resulting in better GHS and body image and less symptoms compared to radiotherapy with or without chemotherapy. There was no difference in nancial di culties by treatment type.

Discussion
This is a maiden study conducted in Ghana to investigate HRQoL among cervical cancer survivors. The GHS/overall QoL was satisfactory. The score in functional domains were high, with highest scores seen in social functioning, followed by role functioning. Financial di culties, pain and fatigue were the most frequently reported symptoms. Participants with FIGO stage II cervical cancer had lower GHS/overall QoL score and greater nancial di culties. Compared to other modes of treatment, surgery was associated with better overall QoL and body image. Chemoradiation was associated with more clinical complaints and nancial di culties. Illiteracy, body image, pain and loss of appetite were signi cantly associated with overall GHS/overall QoL. Cervical cancer and its treatment have signi cant impacts on the HRQoL, and thus its management should not be focused only on overall survival, but rather a holistic care also aimed at maximizing QoL.
Time since completion of treatment has effects on participants' perception of HRQoL. 20 In our study, the pattern of change in GHS was highest at or after 15 years of treatment. The perceived increase in GHS was later when compared to the trend reported in a similar study in China. 20 This difference may be attributed to the mean age of the participants and the capacity to rehabilitate subjects after treatment. Similar recovery patterns were also noticed in all the EORTC functional domains with the exception of physical functioning, which was less brisk. Survivors reported worse score on this HRQoL domain with increasing period of survivorship. 21 The course of GHS appears to have an inverse relationship with the pattern of symptom complaints. Report of good GHS by survivors occur at the time of improved functioning, and they are unlikely to report symptoms during this period. This may explain the low symptom experience score. Comparable studies reported score range of 15.6 to 56 and 15.9 to 47 for and pain and fatigue respectively within a year of treatment completion. 15,17,20,22 These scores compare to the score of 20.6 and 19.4 for fatigue and pain recorded respectively within a year of the treatment in this current study. Further, we noticed an upward trend in complaint of pain within 10 years following treatment. An earlier study by Brown et al., 2014, reported 5 to 10 in 100 cervical cancer survivors complained of pain. 23 Sexual function was a primary source of distress among cancer survivors. 24 It is an issues of paramount importance especially when a signi cant number of survivors are young in this study. At the time of the interview, only 56 (36.6%) were sexually active. This was comparatively higher than the rate reported among survivors in a similar study in Iran. 25 The level of sexual activity among survivor may have sociodemographic determinants, with increased sexual activity in a more liberal society. 26 Higher score in sexual domains is also reported in participants reporting with early disease. Sexual worry (perception that sex will be painful) was higher within the rst year of treatment, but thereafter, and then improved. Actual sexual activity decreased along the period of follow-up, although vaginal function and sexual enjoyment improved. 20 Female survivor populations treated for an early disease tend to have good GHS/overall QoL. 27 In our study, the global health status for the cervical cancer survivors was 79.7 ± 16.2. This complements the ndings of a similar in Taiwan which had few women surviving with stage IV disease. 28 A study that had relatively more women with advanced cervical cancer patient reported with a global health status of 59.5 ± 10.9 (after excluding critically ill patients). 29 The differences in the global health status scores between these studies is related to the differences in treatment options and time since completion of treatment.
Participants treated for FIGO stage I disease reported the highest GHS followed by that of stage III. This nding was consistent with that of an earlier study by Thapa et al., 2018 in Hubei, China. 20 However in a similar work in Malaysia, women treated for stage III reported the highest GHS. 30 No clear relationship exists between GHS and the FIGO stage of cervical cancer. The disparity in perceptions of GHS may be explained by the response shift theory, which postulates that perception of overall health is in uenced by a change in ones' condition of health. 31 Survivors who receive radiation with or without chemotherapy have worse HRQoL. 20,26 Relatively more women who survived FIGO stage II disease had chemoradiation, that may underline the poor perception of GHS in this group of survivors. The actual cost of treatment for stage II or III along with the cost accrued from loss of productivity is high with chemoradiation. 32 Treatment of cervical cancer is primarily by surgery or radiation therapy, with chemotherapy as valuable adjunct. 3 In consonance with earlier studies, survivors who received surgery as a primary treatment modality in this study had lesser impact on their GHS, followed by survivors who received radiotherapy alone. The GHS, body image, nancial di culties and symptoms reported by survivors differed with the treatment option. In this study, surgery was associated with the highest GHS/overall QoL, which is consistent with earlier studies. 17,20 Earlier studies also reported more symptoms in women treated with radiotherapy with or without chemotherapy. 33 The symptoms experienced after treatment may underline the signi cant alteration in body image observed across the various treatment modalities. The loss of reproductive organs and external scarring of the genitalia via radical surgery and radiotherapy respectively, have been reported to negatively impact survivors' psychophysical identity and cause distortion of the body image. 33 The absence of uterus and cervix, and the inability to reproduce traditionally, although often overlooked, is a major source of stress or discomfort to young women living with or surviving cancer. 34 These organs are seen by survivors as embodiment of womanhood, femininity and fertility. In most of the cultures in Sub-Saharan Africa, childbearing is a signi cant expression of femininity.
Prior studies demonstrate that women who had radiotherapy, were worse off nancially, compared to those who received surgery alone or surgery with postoperative radiotherapy (PORT). 17,26 These ndings were mirrored in this current study. Longer treatment time, and the need to suspend work during and after treatment may explain this phenomenon. 35 Minimal use of PORT maximizes the gains after surgery, and multi-modal treatment may unnecessarily overburden the surgical and radiation oncology facilities which are already inadequate in low-resource countries. 36,37 This places emphasis on diligent pre-treatment assessment especially in subjects with operable tumours. MRI should be considered in women being prepared for radical hysterectomy and pelvic lymph node dissection. Although accurate information about tumour size and lymph node involvement can be obtained from surgical pathology, it is desirable to estimate tumour volume and the presence of lymph node metastases or adverse prognostic ndings before intended surgery. Women with node-positive disease or other surgical risk factors could then be managed appropriately, avoiding the potential drop in GHS associated with a combination of radical surgery and post-operative radiotherapy. 38 Disparities may exist in terms of demographic factors for cervical cancer patients and survivors. While cervical cancer patients tend to be less educated and often dwell in rural areas, survivors as reported by this study were better educated and only few had rural residence. 39 Surprisingly, survivors with no formal education were less likely to describe their GHS as "affected". An earlier study on cervical cancer survivors did not report the role of level of education on HRQoL. 20 Pain was a major predictor of GHS among this group of survivors, which is consistent with an earlier study with similar clinical and treatment characteristics. 28 The cause of pain in the survivor may not necessarily be attributable to the cancer or its management. 23 Current interventions in cancer care may afford survivors the chance to live longer and experience pain that may be secondary to the aging process or senescence. Loss of appetite has been reported among subjects during and after treatment. 17,25,40 Subjects who had lost appetite were likely to report an affected GHS in this study. Similar reports were recorded among survivors in Iran and Bangledash. 17,25 From these ndings, good diet and nutrition extend well into the period of remission, and signi cantly in uencing the QoL of the CCS. On body image, feeling less physically attractive, less feminine or been dissatis ed with the self-image were predictive of an affected GHS in this study. The absence of uterus and cervix, and the inability to reproduce traditionally, although often overlooked, is a major source of stress or discomfort to young women living with or surviving cancer. 34 For these survivors, the organs are traditionally seen as embodiment of womanhood, femininity and fertility, and for most of cultures in Sub-Saharan Africa, childbearing is a signi cant expression of femininity.
Limitations of this study include that the EORTC assessment scale has not been validated in Ghana. The closest locally relevant interpretation of questions was adopted by the researchers. The GHS of cancer survivors' changes over time. As this is a cross-sectional design, the assessment of GHS was not done over time, and the lack of the comparison of GHS score before and after treatment contribute to the limitations. Due to the retrospective nature of the treatment data, there is a possibility that some of clinical and treatment information may not have been documented. This was a single institution study so this population of cervical cancer survivors may not be representative of the cervical cancer survivor population in Ghana or in other LMICs. However, the study contributes to how to improve patient care and further research for women with CC in Ghana. Additional longitudinal and intervention studies with control groups may further evaluate the HRQoL of CCS.

Conclusion
Cervical cancer and its treatment affect survivors' HRQoL. Among cervical cancer survivors in Ghana, the HRQoL was satisfactory. Surgery offers the best prospects for HRQoL with the least nancial burden. Efforts to enhance HRQoL should also be focused on maximizing survivors' body image and managing pain and loss of appetite. Educating women with cervical cancer about expected impacts on their quality of life, and strategies to manage and mitigate these impacts, is essential.