Quality of life of patients with cancer undergoing chemotherapy in hospitals in Belo Horizonte, Minas Gerais State, Brazil: does individual characteristics matter? Qualidade de vida de pacientes com câncer submetidos a quimioterapia em hospitais em Belo Horizonte, Minas Gerais, Brasil: as características individuais importam? Calidad de vida de pacientes con cáncer sometidos a quimioterapia en hospitales de Belo Horizonte, Minas Gerais, Brasil: ¿importan las características individuales?

This study aims to evaluate changes in quality of life of cancer patients at the beginning of the first and the second cycle of chemotherapy (CT) in hospitals in Belo Horizonte, Minas Gerais State, Brazil. Longitudinal, prospective, descriptive study with a quantitative approach. We enrolled 230 patients, from a broader cohort, diagnosed with the five most frequent types of cancer (breast, colorectal, cervical, lung, and head and neck), aged 18 years or older, who were initiating CT for the first time. quality of life was assessed with the EORTC QLQ-C30 version 3, applied at the beginning of the first and second CT cycle. The paired Wilcoxon test was used to identify differences in quality of life between the two time points. A multivariate linear regression analysis was performed using the bootstrap method to investigate potential predictors of global health Status/quality of life. There was a significant increase in patients’ emotional function scores (p < 0.001) as well as symptom scores for pain (p = 0.026), diarrhea (p = 0.018), and nausea/vomiting (p < 0.001) after initiation of CT. Widowhood was associated with improvements in the global health Status/quality of life (p = 0.028), whereas the presence of cervical cancer (p = 0.034) and being underweight (p = 0.033) were related to poorer global health status/quality of life scores. CT has detrimental effects on patients’ physical health but, on the other hand, it leads to improvements in the emotional domain. Patients’ individual characteristics at the beginning of CT are associated with changes in their quality of life. Our study could help identify these characteristics. Neoplasms; Quality of Life; Drug Therapy; Health Profile; Cytotoxins Correspondence D. P. Moreira Rua Bernardo Guimarães 341, apto. 102, Belo Horizonte, MG 30140-080, Brasil. danipenamoreira@gmail.com 1 Universidade Federal de Minas Gerais, Belo Horizonte, Brasil. This article is published in Open Access under the Creative Commons Attribution license, which allows use, distribution, and reproduction in any medium, without restrictions, as long as the original work is correctly cited. Cad. Saúde Pública 2021; 37(8):e00002220


Data collection
Sociodemographic data were collected using an instrument developed by the investigators, the Clinical and Sociodemographic Identification Questionnaire (IC), which was administered to subjects on the first day of the first chemotherapy cycle. Clinical data were collected from patient charts. The EORTC QLQ-C30 version 3.0 was at two different timepoints -on the first day of the first chemotherapy cycle (timepoint 1) and on the first day of the second chemotherapy cycle (timepoint 2) -in order to assess the quality of life of patients. The EORTC QLQ-C30 version 3.0 instrument is composed of 30 questions and divided into multiple-item and single-item scales. In total, five functional scales (physical, role, emotional, cognitive and social), nine symptoms/single-item scales (fatigue, nausea/vomiting and pain, dyspnea, insomnia, appetite loss, constipation, diarrhea and financial difficulties) and a global health status/quality of life scale were included. Scales are linearly transformed into scores ranging from 0 to 100. High scores in the global health status/quality of life and functional scales represent high quality of life and a healthy functional level. In symptoms or single-item scales, high scores represent a high level of symptoms or problems 15 . Data were collected from June to November 2015.

Statistical analyses
The descriptive analysis was based on the calculation of absolute and relative frequencies for categorical variables, and measures of central tendency and dispersion (mean, median, standard deviation, and minimum and maximum) for quantitative variables. For each scale of the instrument, patient scores were calculated according to the instructions in the manual 15 . The distribution of scores was not considered normal (Shapiro-Wilk test). The paired Wilcoxon test was used to identify significant differences between EORTC QLQ-C30 timepoint 1 and timepoint 2. A regression model was constructed with GHS/QoL at timepoint 2 as the dependent variable, adjusted by global health status/quality of life timepoint 1, the treating hospital and the stage of cancer at the time of diagnosis. Sociodemographic, clinical, and treatment-related variables were used as the independent ones. For the variable selection, univariate models with global health status/quality of life at timepoint 2 were considered as dependent variable and clinical, sociodemographic or treatment-related as the independent variable. A p-value < 0.2 was used as criteria for this initial selection. Based on this set of pre-selected variables, a final multiple linear regression model was considered based on the backward selection technique. This method initially considers all variables in the model. Adjustments were made by removing variables one by one, and those that did not meet a certain significance level are excluded. This procedure was repeated until all variables were significant at 5% level. The non-parametric bootstrap method was used to calculate confidence intervals and p-values related to the regression model, with 2,000 replications. Also, 95% confidence intervals and a 0.05 level of significance were used. All analyses were performed with R statistical software (http://www.r-project.org).

Compliance with ethical standards
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
The Institutional Review Board at the UFMG, approved the study (CAAE: 36059514.5.0000.5149) and informed consent was obtained from all individual participants included in the study. This article does not contain any studies with animals performed by any of the authors.

The patients
We selected the five most frequent types of cancer in the original cohort (breast, colon and rectum, cervix, head and neck, and lung), which resulted in 230 patients (65.3% of the 315 patients). This choice would make possible perform the statistical analysis in a proper way (Figure 1).

Figure 1
Inclusion flow and follow-up of patients. Belo Horizonte, Minas Gerais State, Brazil, 2020. Demographic data, clinical characteristics and variables related to the treatment of 230 patients according to the type of cancer are described in Table 1. Most study subjects were female, with a mean age of 56.1 years, self-declared mixed-race, and married. The mean time of schooling was 6.2 years, 14.3% had no income, and most (45.2%) did not earn more than one minimum wage (USD 200) -considering the Brazilian minimum wage in 2015 and the USD to Brazilian real exchange rate in December 2015. Regarding the lifestyle of study participants, most were or currently are tobacco and alcohol users. As for clinical characteristics, most subjects had breast cancer classified as stages III and IV at the time of diagnosis. Most patients were not overweight and had one comorbidity. The most prevalent comorbidities in this study were systemic hypertension followed by diabetes mellitus, depression, and cardiovascular diseases. According to the type of cancer, two of them are exclusively female (breast and cervix), and the other types, females represented half of the patients with lung cancer, 54.3% of the colon and rectum and only 15% of the head and neck. The patients with colon and rectal and lung cancer were older, 60.5 and 59.4 years old, respectively, and for all types of cancer, the most participants were mixed-race and married. Regarding years of study, the lowest averages were found for patients with head and neck and cervix cancer, 5.3 years for both, and patients with lung cancer had the highest proportion of complete high school (21.3%). The frequency of patients without income was higher among patients with cervical cancer (31.8%) and the highest proportion of patients who received more than two minimum wages was those with lung cancer (31.3%). More than half of the patients with head and neck and lung cancer were smokers, 67.5% and 65.6%, respectively, and 92.5% of patients with head and neck cancer were alcoholics. In all types of cancer, patients were diagnosed more frequently in stages III and IV and only for breast cancer the majority were obese (44.1%). Most patients with cervical and head and neck cancer did not have any comorbidity, 40.9% and 47.5%, respectively.

CT
More than a half of the patients had previously undergone a different type of treatment, and the mean time between diagnosis and the first treatment was 50.6 days. The most frequent chemotherapy protocol in this study was cisplatin (CDDP). The mean distance traveled by subjects for chemotherapy treatments was 95.9 kilometers (km) and 34.3% were residents of Belo Horizonte. Although 20.9% of the sample stated to have private health insurance, most chemotherapy treatments were covered by the Brazilian Unified National Health System (SUS). Most patients with cervix, head and neck and lung cancer underwent treatment prior to CT, 61.4%, 62.5% and 71.9%, respectively. Patients with cervical cancer presented the highest average number of days to start treatment (63.7 days) and 50% of these patients started treatment after 60 days. Furthermore, they traveled long distances from the municipality they dwell to the treatment municipality, with an average of 117km. Lung cancer patients presented the longest distance traveled, , with an average of 119km.

Quality of life
The Wilcoxon paired test identified virtually no changes in the global health status/quality of life score of patients between timepoints 1 and 2, and the scores were considered high. In the median score for emotional function a significant improvement at timepoint 2 (p < 0.001) was identified. Regarding symptoms, a significant increase in the median scores for nausea and vomiting (p < 0.001) and diarrhea (p = 0.018) was detected, and a significant drop in the median scores for pain (p = 0.026) at timepoint 2. No significant difference between the timepoints on other scales ( Table 2) was observed.
When looking at the association between clinical and sociodemographic variables, we observed that marital status, type of cancer, and BMI were predictors of changes in global health status/quality of life. After the initiation of treatment, patients with cervical cancer reported deterioration of their global health status/quality of life when compared to those with breast cancer (β = -7.97, p = 0.034) as well as underweight patients compared with those in their normal weight range (β = -9.03, p = 0.033). Widowed subjects reported improvement of their global health status/quality of life score at timepoint 2 when compared to married subjects (β = 6.9, p = 0.028) ( Table 3).

Discussion
This study analyzed the quality of life of cancer patients with different types of cancer undergoing CT in hospitals in Belo Horizonte. We observed that in the functional quality of life scales, only the emotional domain showed a statistically significant improvement when comparing both treatment cycles. Significant changes in some symptom scales were identified. Pain, nausea/vomiting and diarrhea worsened after starting treatment. Regarding the global health status/quality of life, after starting treatment, patients with cervical cancer reported worsening, as well as patients with low weight. Widow/widower reported improvement in global health status/quality of life at timepoint 2.
In this and other studies 17,18 , a significant increase in emotional function scores were identified after beginning chemotherapy treatment. The improvement can be related to the anxiety experienced by patients while waiting for CT treatment initiation. After acceptance of the diagnosis, treatment is initiated aiming disease cure, which increases patients' level of confidence. Also, treatment helps patients regain hope of improvement 19 . Cancer is a stigmatizing disease, which is associated with the possibility of suffering and death. Finally, getting treated may bring comfort, improving patients' emotional status 20 . Chemotherapeutic agents act at different stages of the cell cycle to inhibit mitosis and induce cell death. Due to their non-specific action, these drugs also affect healthy cells in the body, which explains the need for intervals between CT treatments to give cells time to recover 21 . The adverse effects of CT are primarily explained by the non-specific action of chemotherapeutic agents, and some of the most frequent are nausea and vomiting 22 . Chemotherapy-induced nausea and vomiting are common adverse effects in patients undergoing CT, and this may discourage adherence to treatment, as well as negatively affect quality of life and clinical status 23,24 .
Chemotherapeutic agents are classified according to their potential to cause nausea and vomiting, in order to define guidelines for antiemetic management. This classification is divided into high,  moderate, low, and minimal emetic potential. The two most frequent protocols in this study are classified as highly likely to cause nausea and vomiting 25 . Despite improvements over the years in different countries, guidelines for antiemetic management still show very poor adherence levels 26 . In Brazil, the SUS offers different classes of antiemetic drugs, free of cost, such as serotonin receptor antagonists (ondansetron), dopamine antagonists (metoclopramide), among others 27 . Ondansetron is an antiemetic agent effective for patients undergoing treatment with high and medium-emeticpotential chemotherapeutic drugs 28 , although no protocols and therapeutic guidelines are available in Brazil for the prevention of chemotherapy-induced nausea and vomiting. Similarly to this study, we have found other studies that detected an increase in nausea and vomiting in patients undergoing CT 17,29 .
Cad. Saúde Pública 2021; 37(8):e00002220 Diarrhea also increased significantly after chemotherapy treatment. CT-induced diarrhea can occur due to changes in intestinal absorption caused by chemotherapeutic agents, in addition to biochemical and microbial changes 30 . Other studies have also reported increased diarrhea after CT 11,13,31 .
A systematic review 32 showed that CT improves pain control, supporting our findings indicating that pain considerably decreased after CT. The main targets of CT are: control of the disease, mass shrinking, cure, and palliative care. Despite the growing number of studies on pain, the better understanding of its mechanisms and the increase in the consumption of drugs to relieve the suffering, it was found in another systematic review that the prevalence of pain among cancer patients remains high. A possible explanation for this is the change in the profile of cancer patients, which are living longer and are more likely to report their pain 33 . Pain may be caused by invasion or compression of adjacent structures by the tumor and may be related to adverse effects of cancer treatments 34 . Identifying pain in cancer patients and treating it is extremely important due to the impact on their quality of life 35 .
In this study, widowhood was a predictor of changes in the global health status/quality of life after beginning CT, when compared to married subjects. Older age is associated with widowed status, and older adults experience fewer negative emotions than younger people. Besides, when these emotions are experienced, they are also not as intense (for the older adults) 36 . As time passes and people age, they realize time is finite and gain a deeper appreciation for emotional meanings 37 , which may explain the finding that a widowed status was associated with improvements in the global health status/quality of life after CT. Another possible explanation is the attention received from healthcare professionals during disease. Studies have shown that widowed subjects report feeling more lonely than married subjects 38,39 , therefore contact with the team involved in their care may be related to a positive self-perception of quality of life in this group of patients.
The type of cancer was also a predictor of changes in the global health status/quality of life after beginning the CT treatment. The presence of cervical cancer is associated with quality of life deterioration after chemotherapy, compared to breast cancer. The type of treatment for cervical cancer depends on the stage of the disease, and also on the clinical conditions of the patient. Most women treated with CT are also undergoing concomitant radiation therapy 40 . Several factors can affect the quality of life in women undergoing treatment for cervical cancer, including changes in sexual functioning due to vaginal stenosis, decreased libido, and pain during intercourse, in addition to symptoms such as lymphedema and urinary and fecal incontinence 41 . A relevant point to be discussed is the relationship between this type of cancer and poverty 42 . In this study, a significant difference in individual income and years of schooling was found between women with breast cancer and cervical cancer, with lower means in the latter group. Some studies have found a negative association between decreased quality of life and low income among cancer patients 43,44,45 .
Body mass index (BMI) is an international indicator of the nutritional status of individuals. The cutoff points used for this calculation follow the recommendation of the WHO, namely: < 18.5kg/ m² (underweight), > 18.5kg/m² and < 25kg/m² (ideal weight range), > 25kg/m² and < 30kg/m² (overweight) and > 30kg/m² (obese) 46 . Cancer-associated cachexia is a multifactorial disorder, characterized mainly by loss of body weight resulting not only from adipose tissue but also from muscle mass. There is still no medical intervention capable of reversing the nutritional support of patients affected by cachexia, and this disorder compromises patients' physical and emotional functions as well as their quality of life 47 . One of the classifications for cachexia is a BMI < 20kg/m² 48 . This syndrome is associated with deterioration of patient's quality of life 49,50 , confirming the findings of this study, in which underweight patients reported poorer global health status/quality of life after the onset of treatment, when compared to those in their normal weight range. One explanation for this deteriorated quality of life is the increased toxicity inherent to the treatment, due to loss of muscle 47 .
Similar to another study 51 , it draws attention to the large number of patients who did not start treatment within 60 days, as advocated by Brazilian regulation 52 that guarantees the right of patients with malignant neoplasia to begin their first treatment in SUS within a maximum period of 60 days after diagnosis. Although this study did not find an association of changes in the quality of lifeL of patients with the interval between diagnosis and the start of treatment, the delay in beginning treat-Cad. Saúde Pública 2021; 37(8):e00002220 ment can negatively affect the patients' survival 53,54 and the agility to beginning treatment improves quality of life 54 .
One of the limitations of this study is the poor representativeness of some high-incidence cancers that use CT in a small portion of the population, such as prostate cancer, for example. Also, we collected data at the beginning of CT, and therefore could not assess the variability of quality of life predictors over the entire course of treatment. Another significant limitation of this study is the loss of patient data due to incomplete hospital records and absence of analysis for each type of cancer separately.

Conclusion
The adverse effects of CT identified in this study, such as nausea, vomiting, and diarrhea, have major impacts on the quality of life of cancer patients. Despite its physical deterioration consequences, we also observed that (chemo)therapy had a positive effect on patients' emotional state which may contribute and arouse future discussions. Furthermore, we observed that individual characteristics are associated with changes in the global health status/quality of life of patients after treatment initiation. Assessing the quality of life of patients with different types of cancer on CT using their clinical and individual characteristics may help identify those who are more vulnerable to quality of life deterioration and provide information to support decisions made by healthcare professionals who deal with the care of patients with different neoplasms. With the growing advances in diagnosis and treatment of malignant tumors, we are now aiming to change prevalence and provide longer survivals. Therefore, it is greatly significant to consider the change in the comfort level expected by cancer patients. Studies looking at the quality of life of patients with cancer are important to generate information for healthcare professionals to ensure patients a dignified and functional life by monitoring signs and symptoms not only inherent to the disease, but also arising from therapy and patients' individual characteristics.